Universality class of replica symmetry breaking, scaling behavior, and the low-temperature

合集下载

可变区间混沌映射在图像加密中的应用研究

可变区间混沌映射在图像加密中的应用研究

实验 结果表 明 , 该算 法是 一个具有 较好 的安全性 能和抗 攻击 能力的混沌 加 密算法 。
关 键 词 : 段 线 性 混 沌 映 射 ; 标 准 映 射 ; 混 沌 密码 ; 密 码 分 析 ; 序 列 密码 分 中 图 法 分 类 号 : P 0 +7 T 39 . 文献标 识码 : A 文 章 编 号 :0 072 2 1) 45 7.3 10 .04(0 0 2 。160
0 引 言
传 统 加 密 技 术 通 常把 多媒 体信 息作 为普 通 数 据 流 进 行 加 密, 因此 存 在 一 定 的 局 限性 。目前 , 究 人 员 已经 提 出 了很 多 研 基 于 离 散 混 沌 映 射 的 加 密 系 统 , 应 地 也 出现 了 大 量 分 析 混 相 沌 密 码 系 统 的 方 法 “ 。混 沌 具 有 如 下 的一 些 特 征 使 之 很 适 合 捌
57 16
2 1,1(4 00 3 2)
计 算 机 工 程 与 设 计 C mpt nier g D g
O信 息安 全 技 术 ・
可变区间混沌映射在图像加 密中的应 用研究
李 波 ( 重庆教 育学 院 计 算机 与现代教 育技 术 系,重 庆 4 0 6) 00 7
1 混 沌 映射 的选 择
11 伪 随机 序 列 的生 成 .
为 了充 分 利 用 分 段 线 性 混 沌 映 射 的运 算 速 度 快 的 特 点 ,
本 文 采 用 文 献 [] 出 的 区 间 数 目参 数 化 的 P C 来 产 生 拟 混 3提 LM
沌 伪 随机 序 列 。 同分 段 区 间 P C 相 比较 , 混 沌 映 射 具 有 LM 该 更 高 的 安 全 性 。区 间数 目参 数 化 P C 的 数 学 表 达 式 为 LM

Universality for eigenvalue correlations at the origin of the spectrum

Universality for eigenvalue correlations at the origin of the spectrum
n−1
Kn (x, y ) =
wn (x)
wn (y )
j =0
pj,n(x)pj,n (y ) γn−1,n pn,n (x)pn−1,n (y ) − pn−1,n (x)pn,n (y ) , γn,n x−y
=
wn (x)
wn (y )
(1.3)
which gives the connection with orthogonal polynomials. The second equality in (1.3) follows from the Christoffel-Darboux formula [35]. Akemann et al. [2] showed that the local eigenvalue correlations at the origin of the spectrum have a universal behavior, described in terms of the following Bessel kernel (πu)Jα− 1 (πv ) − Jα− 1 (πu)Jα+ 1 (πv ) √ √ Jα+ 1 2 2 2 2 , Jo α (u, v ) = π u v 2(u − v ) (1.4)
The matrices M are n × n Hermitian and dM is the associated flat Lebesgue measure on the space of n × n Hermitian matrices, and Zn is a normalizing constant (partition function). The confining potential V in (1.1) is a real valued function with enough increase at infinity, for example a polynomial of even degree with positive leading coefficient. Random matrix ensembles are important in many branches of mathematics and physics, see the recent survey paper [18]. The specific ensemble (1.1) is relevant in three-dimensional quantum chromodynamics [39]. The ensemble (1.1) induces a probability density function on the n eigenvalues x1 , . . . , xn of M , given by P (n) (x1 , . . . , xn ) =

基于加权奇异值分解截断共轭梯度的电容层析图像重建

基于加权奇异值分解截断共轭梯度的电容层析图像重建

基于加权奇异值分解截断共轭梯度的电容层析图像重建陈宇;高宝庆;张立新;陈德运;于晓洋【期刊名称】《光学精密工程》【年(卷),期】2010(018)003【摘要】针对电容层析成像技术(ECT)中的"软场"效应和病态问题,提出了一种基于加权奇异值分解(SVD)截断共轭梯度的电容层析(ECT)图像重建算法.阐述了电容层析成像工作原理,提出了12电极ECT系统的测量方法.在分析灵敏度矩阵的奇异值分解理论的基础上,推导出了加权SVD截断共轭梯度的数学模型,并利用Tikhonov 方法进行正则化加权处理.最后,分析了算法的收敛性,并将其应用于电容层析成像系统的图像重建中.实验结果表明,对于层流,截断共轭梯度算法的平均误差能达到27.54%,全部流型平均迭代步数达到13步,与LBP、Landweber和CG算法比较,该算法具有成像效果好,成像速度快,易于实现等特点.【总页数】7页(P701-707)【作者】陈宇;高宝庆;张立新;陈德运;于晓洋【作者单位】哈尔滨理工大学,黑龙江,哈尔滨,150080;东北林业大学,黑龙江,哈尔滨,150040;哈尔滨理工大学,黑龙江,哈尔滨,150080;哈尔滨理工大学,黑龙江,哈尔滨,150080;哈尔滨理工大学,黑龙江,哈尔滨,150080;哈尔滨理工大学,黑龙江,哈尔滨,150080【正文语种】中文【中图分类】TP391.4【相关文献】1.一种基于截断奇异值分解正则化的电离层层析成像算法 [J], 欧明;甄卫民;於晓;徐继生;邓忠新2.三项共轭梯度的电容层析成像图像重建算法 [J], 陈宇;孙帆;张健3.基于数据驱动的卷积神经网络电容层析成像图像重建 [J], 孙先亮; 李健; 韩哲哲; 许传龙4.基于改进ALEXNET卷积神经网络的电容层析成像三维图像重建 [J], 李岩;王璐;李佳琪5.基于概率加权共轭梯度算法的混凝土超声波层析成像 [J], 刘建军;许令周因版权原因,仅展示原文概要,查看原文内容请购买。

FullProf 详细使用教程

FullProf 详细使用教程
Durban, August 24, 2003 ECM-21 Software Workshop
New facilities concerning symmetry in FullProf
⇒ The symmetry used within FullProf is totally based in the Crystallographic Fortran 95 Modules Library (CrysFML) (Tuesday 26 ⇒ FA3-MS5, Meeting Room 11B) ⇒ These modules provide better crystallographic information to the user of the program. In particular automatic calculation of the multiplicity of each site is now performed after reading the atoms as well as the calculation of the appropriate coefficients for automatic quantitative analysis of mixture of phases. ⇒ New output files with full information of crystallographic symmetry are produced (extension: sym)
Durban, August 24, 2003
ECM-21 Software Workshop
Last minute changes in FullProf Documented in “fp2k.inf”

利用有限域上奇异辛几何构造一个新的带仲裁的认证码

利用有限域上奇异辛几何构造一个新的带仲裁的认证码


) .
收稿 日 :0 00—5 作者简介:高有 (96 月生) 期 2 1—30 . 16 年7 ,男,博 士,教授. 究方 向:代数 、编码与密码 研 基金项 目:国家 自 然科学基金 (1 706 天津 自 6 192 ) 然科学基金 (8 c J 30 ) OJ YB c19 0.
分 别 称 S E , 和 为 信 源 集 ,发 方 编 码 规 则集 , 收 方 解 码 规 则 集 和信 息 集 , 集 , T ER
合 fIlT,ER, I S,E II Il 的基数称 为这个认证码 的参数 .在 一个具有仲裁 的认证系统 中,有 4 个
参与者 :发方 ,收方 ,敌方和仲裁人 ,共有下面五种攻击:

ma {, 一 一s x 0m )
mi{, 一2} n/m 8,
成立 .若 ( s尼2 m,,; +z 是非空 的,那么它 在 2 ,(q作用 下构成一子空 问轨道 . 由 , ) 蚪l F )
文献 [ ] 1 知 6
M ( s ;u ,) =g。+一 +)(一)一) m,, 2 +e l 2 。m k m 知 ( + (
1 敌方 的模 仿攻击 .敌方在未观 测到任何信 息的条件 下 ,发送 一个信息给 收方 ,若收 方 ) 将其 当作合法信 息接 收 ( 即该信 息满足 收方 的认证 条件) ,则称敌方 的模仿攻击成 功.假定编 码 规则服从均匀分布,将敌方模仿攻击成功的最大概率记 为 P, ,则

m ax



( P P 合 同于 i ) l
, 、
c s , =
( 。。 一8 。) s ; 一
- 4 -
记F
的所有的 ( s ) m,, 型子空间所成的集合 为M ( s ;u- 1 ) m,, 2 4 , .由文献 【 】 - 1 可 6

外文翻译---基于离散混沌映射的图像加密并行算法

外文翻译---基于离散混沌映射的图像加密并行算法
这种并行图像加密框架下,我们提出了一种新的算法,这是基于四个基本的转换。因此,我们将描述我们的算法之前,先介绍这些转换。
3.转换
3.1.A-转换
在A转换中,A代表加,能被形式化的定义如下:
a+b=c(1)
加法被定义为按位与操作
转换A有三个基本性质:
(2.1)a+a=0
(2.2)a+b=b+a(2)
(2.3)(a+b)+c=a+(b+c)
在并行模式计算时,许多的PE可以同时读取或写入相同的内存区域(即临界区),
这往往会导致意想不到的执行程序。因此,有必要在关键区域使用一些并行技术管理。
2.2.并行图像的加密框架
为了满足上述要求,我们提出了一个并行图像加密的框架,这是一个四个步骤的过程:
步骤1:整个图像被划分成若干块。
步骤2:每个PE负责确定数量块。一个区域内的像素可以充分使用有效的混乱和扩散进行操作加密。
附件C:译文
基于离散混沌映射的图像加密并行算法
摘要:
最近,针对图像加密提出了多种基于混沌的算法。然而,它们都无法在并行计算环境中有效工作。在本文中,我们提出了一个并行图像加密的框架。基于此框架内,一个使用离散柯尔莫哥洛夫流映射的新算法被提出。它符合所有并行图像加密算法的要求。此外,它是安全、快速的。这些特性使得它是一个很好的基于并行计算平台上的图像加密选择。
这个框架可以非常有效的实现整个图像的扩散。但是,它是不适合在并行计算环境中运行。这是因为当前像素的处理无法启动直到前一个像素已加密。即使有多个处理元素(PE),这种计算仍然是在一个串行模式下工作。此限制了其应用平台,因为许多基于FPGA / CPLD或者数字电路的设备可以支持并行处理。随着并行计算技术的应用,加密速度可以大大加快。

利用奇异酉几何构造新的带仲裁的认证码

利用奇异酉几何构造新的带仲裁的认证码
功 的概率分别记为 , s P, , 。 和 . 仲裁人 了解通信系统的全部但 不参与通信活动 , 只有 当发方与收方发生争执时, 才出面解 决争端 , 因此假定仲裁者总是诚实的. 文献 『 利用非奇异 辛几 何构造 了一个无仲 裁的认证码, 献f 6利用 非奇异酉几何 、非奇 2 1 文 31 — 异辛几何构造 了一系列带仲裁 的认证码, 文献f 利用奇异辛几何构造 了一个带仲 裁的认证码, 7 1 这 些结果极大地丰 富了有限典型群的几何学在认证码构造 方面的应用. 本文 利用奇异酉几何构造 了一个新的带仲裁 的认证码, 并计算 了这个码的参数与各种 攻击成功 的概率. 所有酉几何中的符号及计数公式见文献f . 8 1
设F。 日 是具有q个元素的有限域, 其中q 是素数幂. 定义F: q 的对合自同构, 面=a, 0 q 任意 取n q. ∈F 它是一个二阶自同构, 其不变域是F . q

I:I ) 0 l
/ 0

0) (/ 1
I .
上满足 = 的所有( +f×( +/ 2 ) 2 ) u 阶非奇异矩阵构成一个群, 称为F。:v 级, q 2 +f J 秩为2 的奇异酉群, 用 外1uF。 , (a 来表示. 2 ) 设 蚪 q 是F。 上的( 2 +z ) 维向量空间. 定义 外 (q 在 蚪 F。 ) 上的作用如下:
证 1 )由sc U上 任取s∈S 可 设 , ,
/ ‘ 0 0 0 0 0 0、 , I 1 R 0 2 0 R 0 1 R 0 3l 0 0 0 0 0 , 0 1 J
\ 0 0 0 0 0 Ik 1 0 / ( ) .
并 定义:
加密函数.: 厂 S×E M, ,T H =s T T ( e) m 8 +e ; 解密函数g: ×E M R u{ 欺诈) , ( e) m ̄R H 8 , { 欺诈) ,

基于辛几何构作一类新的带仲裁的认证码

基于辛几何构作一类新的带仲裁的认证码

i e s n to y te t n mi e ,a s c e su mp ro ai n b h e e v ra d a s c e s ls b t u in b mp ro ai n b h r s t r u c sf li e s n t y t e r c i e n u c s f u si t y a t o u t o
m2
其 中
的秩 为 2。 r 从而 P是一 个含 于 并 且满 足
中所 有元 素正交 的 向量 组成 的集合 , 则
( v) 2
与 的 和 为 ( r 型 子 空 间 的 ( m,) m—m. ) 子 空 一m , 型 r
P ∈ =
,1 、
K =0 V 期
中 国 民 航 大 学 学 报
J oURNAL VI AVI OF CI L ATI oN UNI VEI I Y T oF CHI NA
V0 . 9 12 No 1 .
21 0 1年 2月
F b ay 2 1 e r r 01 u
基 于 辛几 何 构 作 一 类新 的 带仲 裁 的认 证 码
显 然 , 知 ) G 勘 F ) 一 个 子 群 , 定 义 ( 是 L ( q的 则

( ) 在 驯上 的作 用如 下


x S Fq— p ) Fq
基 金 项 目 : 津 市 自然 科 学 基 金 项 目 ( 8 G J 1 9 0) 中 国 民航 大 学 科 研 项 目( 9 A C S 2 天 0 J YB G 3 0 ; 0C U — 0 )
称 为 维 非迷 向子空 间 。
设 P是 F 中的 m 维子 空 间 ,用 P 表 示 由与 P

基于双重混沌映射的图像加密算法

基于双重混沌映射的图像加密算法

组 伪随 机序 列 ,然 后对 L g si 密后 的结 果通 过 改变 图 o i c加 t 像 各点 的像 素 值进 行 二次 加 密 。通 过 两次 加 密使 得加 密效 果 更好 ,安全性 更 高 。
2基础知识
2 1L g t . oii s c混沌系统
Lg t 映射 对 图像 的像 素 进 行 置 乱 ,达 到 改 变图像 像 素 点 的 位 置 ,再 运 用 Lrn 混 沌 系统 对 Lg t 加 密后 的 图像 进 行 二 次加 密 oii sc oez oii sc
来 改 变图像 的像 素 值 以 达 到 双 重 加 密 的效 果 。仿 真 实验 表 明 ,该 算 法 具 有 可 行 性 、执 行 速 度 快 、密钥 量 大和 加 密效 果 好 的优 点 。
基于双重混 沌映射 的图像 加ቤተ መጻሕፍቲ ባይዱ算法
刘剑 鸣 ,李 晓 君
( 山大 学 工业 计算 机控 制工 程 河北 省重 点实 验室 ,河北 秦 皇岛 06 0 ) 燕 6 04
摘 要 :提 出 了一种双 重混沌的图像 加 密算法 。算 法采 用 L gsi 混沌映射和 L o itc rn oe z混沌 系统对图像 进行加 密。算 法首先利 用
关键 词 :混沌 ;图像加 密;置乱 ;l ii映射 ;lez o sc gt o n 混沌 系统 r
Ba e on s d Dua Ch os l a Ma i g ma En yp i Al r t m pp n I ge cr ton go i h
L i —mn L i —jn I Jn i U a g, l×a u o
Ke w o d y r s: ca s i g ecypin srmbig;o it ma pn lrn s se h o ;ma e nr t ;c a l l si o n g c p i oe z y tm g;

基于压缩感知与小波域奇异值分解的图像认证

基于压缩感知与小波域奇异值分解的图像认证

基于压缩感知与小波域奇异值分解的图像认证
高志荣;吕进
【期刊名称】《中南民族大学学报(自然科学版)》
【年(卷),期】2010(029)004
【摘要】提出了一种基于压缩感知与小波域奇异值分解的图像哈希认证新方法.该方法首先通过小波变换得到图像的低频子带,然后对低频子带进行基于块的奇异值分解,最后对产生的最大奇异值数组进行基于压缩感知的随机投影,产生哈希认证信息,并偕同原图像传送到接收端.图像接收端通过比较哈希信息完成图像认证,并通过利用传递的哈希信息重建原始图像的奇异值,进一步映射奇异值差值到图像像素域,从而实现对图像篡改的识别与定位.大量实验结果证明了该方法的有效性.
【总页数】5页(P89-93)
【作者】高志荣;吕进
【作者单位】中南民族大学,计算机科学学院,武汉,430074;中南民族大学,计算机科学学院,武汉,430074
【正文语种】中文
【中图分类】TP309.2
【相关文献】
1.基于小波域奇异值分解的振动信号压缩算法 [J], 王怀光;张培林;陈林;陈彦龙
2.基于奇异值分解的自嵌入图像认证水印算法 [J], 胡玉平
3.一种基于压缩感知的无损图像认证算法 [J], 艾鸽;伍家松;段宇平;舒华忠
4.基于压缩感知的数字图像认证算法研究 [J], 邓桂兵;周爽;李珊珊;蒋天发
5.基于奇异值分解的小波域数字水印方法 [J], 曾晴;马苗;孙莉;周涛
因版权原因,仅展示原文概要,查看原文内容请购买。

Duality symmetry, strong coupling expansion and universal critical amplitudes in two-dimens

Duality symmetry, strong coupling expansion and universal critical amplitudes in two-dimens

a r X iv:c ond-ma t/99868v1[c ond-m at.stat-m ech]4Au g1999Duality symmetry,strong coupling expansion and universal critical amplitudes in two-dimensional Φ4field models Giancarlo Jug INFM –UdR Milano,and Dipartimento di Scienze,Universit`a dell’Insubria Via Lucini 3,22100Como (Italy)∗and Max-Planck-Institut f¨u r Physik komplexer Systeme,N¨o thnitzer Str.38D-01187Dresden (Germany)Boris N.Shalaev Fachbereich Physik,Universit¨a t-Gesamthochschule Essen,D-45117Essen (Germany)∗∗Abstract We show that the exact beta-function β(g )in the continuous 2D g Φ4model possesses the Kramers-Wannier duality symmetry.The duality symmetry transformation ˜g =d (g )such that β(d (g ))=d ′(g )β(g )is constructed and the approximate values of g ∗computed from the duality equation d (g ∗)=g ∗are shown to agree with the available numerical results.The calculation of the beta-function β(g )for the 2D scalar g Φ4field theory based on the strong coupling expansion is developed and the expansion of β(g )in powers of g −1is obtained up to order g −8.The numerical values calculated for the renormalized coupling constant g ∗+are in reasonable good agreement with the best modern estimates recently obtained from the high-temperature series expansion and with those known from the perturbative four-loop renormalization-group calculations.The application of Cardy’s theorem for calculating the renormalized isothermal coupling constant g c of the 2D Ising model and the related univer-sal critical amplitudes is also discussed.PACS numbers:05.50.+q,03.70.+k,64.60.-i,75.10.HkTypeset using REVT E XI.INTRODUCTIONIn this paper we study mainly the symmetry properties of the beta-functionβ(g)for the 2D gΦ4theory,regarded as a continuum limit of the exactly solvable2D Ising model.In contrast to the latter,the2D gΦ4theory is not an integrable quantumfield theory.This means,in particular,that the theory does not possess the factorized scattering matrix,and therefore that the thermodynamic Bethe ansatz method cannot be applied at all.Thus,despite the fact that the2D Ising model at h=0can be solved by many different methods(see[1]for an excellent review),the beta-functionβ(g)of its continuum limit is to date known only in the four-loop approximation within the framework of conventional perturbation theory atfixed dimension d=2[2–4].Calculations of beta-functions are of great interest in statistical mechanics and quantumfield theory.The beta-function contains the essential information on the renormalized coupling constant g∗,this being important for constructing the equation of state of the2D Ising model–for example–which remains still a challenging problem,rich in applications.This and other considerations do not allow us to regard the2D Ising model as having fully been solved.The2D Ising model and some other lattice spin models are known to possess the remark-able Kramers-Wannier(KW)duality symmetry,playing an important role both in statistical mechanics and in quantumfield theory[5–7].The self-duality of the isotropic2D Ising model means that there exists an exact mapping between the high-T and low-T expansions of the partition function[7].In the transfer-matrix language this implies that the transfer-matrix of the model under discussion is covariant under the duality transformation.If we assume that the critical point is unique,the KW self-duality would yield the exact Curie tempera-ture of the model.This holds for a large set of lattice spin models including systems with quenched disorder(for a review see[7,8]).Over twenty years ago the KW self-duality was shown to be equivalent to a Fourier tranformation in target space[9].Also,it has been recognised long ago that self-duality combined with some special algebraic properties of a model leads to the existence of an infinite set of conserved charges[10].Duality is thus known to impose some important constraints on the exact beta-function[11,12].The other main purpose of this paper is to develop a strong coupling expansion for the calculation of the beta-function of the2D scalar gΦ4theory as an alternative approach to standard perturbation theory.It will then be of interest to match this expansion with the results of a four-loop approximation(where possible)by constructing a smooth interpolation with respect to g.It is in fact well known from quantumfield theory and statistical mechanics that any strong coupling expansion is closely connected with a suitable high-temperature (HT)series expansion for a lattice model[1,7].From thefield-theoretical point of view the HT series are nothing but strong coupling expansions forfield models,the lattice being considered as a technical device to define cutoff-regularisedfield theories.Recently,the high-temperature(HT)series expansions and perturbative calculations for the gΦ4field theory atfixed dimensions d<4have been a topic of intense studies(for references see below).Computing critical exponents and various critical amplitude ratios from series expansion data has a long history going back to the early1960s.Nowadays there are a good number of papers containing a large body of information for the N-vector model defined on different lattices for d=2,3,4,5and arbitrary N[13–15].It is remarkable thatthe HT series data for the zerofield susceptibilityχand the second correlation momentµ2 of the N-component classical Heisenberg ferromagnet have been extended up to the order K21(K=J/T),the data for the secondfield derivative of the susceptibility(χ4)being available through to the order K17.Having been equiped with this information,one may try to employ different techniques of resummation of the existing HT series expansions, like Pad´e approximants or more subtle approaches,for computing critical exponents and universal critical amplitude ratios[13,16–21].It is worth noting that the strong-coupling behavior of the gΦ4theory has recently been treated within the framework of a variational perturbative approach[22].The paper is organized as follows.In Sect.II we set up basic notations and define the duality symmetry transformation˜g=d(g).Then it is proved thatβ(d(g))=d′(g)β(g).An approximate expression for d(g)providing good estimates for g∗+(the renormalisedfixed-point coupling constant along the isochore line)is found.In Sect.III the HT series expansion data are used to obtain the strong coupling expansion ofβ(g)for the2D0(N)-symmetric gΦ4 theory in powers of1/g up to the order g−8.Some numerical estimates for the renormalized coupling constant g∗+above T c are obtained.We then compare thefixed point values found to those already known from the four-loop renormalization-group(RG)calculations and from the HT series expansions.In Sect.IV we also discuss the application of Cardy’s formula both for the exact calculation of the renormalized isothermal coupling constant g∗c at T c and,for some universal critical amplitudes,along the isothermal critical line.Sect.V finally contains some concluding remarks.The Appendix presents a simple derivation of the correlation lengthξand of the exact beta-functionβIsing(T)for the lattice2D Ising model, where the temperature T plays the role of an effective coupling constant,and we discuss some of their properties.II.DUALITY SYMMETRY OF THE BETA-FUNCTION We begin by considering the classical Hamiltonian of the2D Ising model(in the absence of an external magneticfield),defined on a square lattice with periodic boundary conditions; as usual:H=−J <i,j>σiσj(2.1)where<i,j>indicates that the summation is over all nearest-neighboring sites;σi=±1are spin variables and J is a spin coupling.The standard definition of the spin-pair correlation function reads:G(R)=<σRσ0>(2.2) where<...>stands for a thermal average.The correlation length may be defined in many different ways,all definitions being equiv-alent to each other in the close vicinity of the critical point[13].This,in fact,reflects the arbitrariness somewhat inherent in any renormalization scheme.The statistical mechanics definition of the correlation length is given by[23]ξ2=d ln G(p)2dµ0(2.5)where d is the spatial dimension(in our case d=2).It should be mentioned that the above definition ofξdiffers from the one used in other related approaches,e.g.[11].The2D Ising model near T c is known to be equivalent to the gΦ4theory with a one-component real order parameter.In order to extend the KW duality symmetry to the continuousfield theory we have need for a”lattice”model definition of the coupling constant g,equivalent to the conventional one exploited in the RG approach.The renormalization coupling constant g of the gΦ4theory is closely related to the fourth derivative of the ”Helmholtz free energy”,namely∂4F(T,m)/∂m4,with respect to the order parameter m= Φ .It may be defined as follows(see[13,14,24]and references therein)g(T,h)=−(∂2χ/∂h2)χ3ξd(2.6)whereχis the homogeneous magnetic susceptibilityχ= d2xG(x)(2.7) It is in fact easy to show that g(T,h)in Eq.(2.6)is merely the standard four-spin correlation function taken at zero external momenta.The renormalized coupling constant of the critical theory is defined by the double limitg∗=limh→0limT→T cg(T,h)(2.8)and it is well known that these limits do not commute with each other.As a result,g∗is a path-dependent quantity in the thermodynamic(T,h)plane[13].Here we are mainly concerned with the coupling constant on the isochore line g(T> T c,h=0)in the disordered phase and with its critical valueg∗+=limT→T+c g(T,h=0)=−∂2χ/∂h2The”lattice”coupling constant g∗+defined in Eq.(2.9)is in a given correspondence with the temperature T c.We shall see that it will be more convenient to deal with a new variable s=exp(2K)tanh(K),where K=J/T.The standard KW duality tranformation is known to be as follows[6,7]sinh(2˜K)=1(1−s)2is a self-dual quantity.Now,on the one hand,we have the formal relationξds(g)dgβ(g)(2.11)where s(g)is defined as the inverse function of g(s),i.e.g(s(g))=g and the beta-function is given,as usual,byξdgdξ=2s(1−s)(1+s(g))(ds(g)/dg)(2.14) Let us define the dual coupling constant˜g and the duality transformation function d(g)ass(˜g)=1s(g))(2.15)where s−1(x)stands for the inverse function of x=s(g).It is easy to check that a further application of the duality map d(g)gives back the original coupling constant,i.e.d(d(g))= g,as it should be.Notice also that the definition of the duality transformation given by Eq.(2.15)has a form similiar to the standard KW duality equation,Eq.(2.10).It is easy to prove that d′(g∗)=±1.The maps we are looking for have d′(g∗)=−1,since the opposite sign leads to the trivial solution d(g)≡g.This is also shown in the Appendix.Consider now the symmetry properties ofβ(g).We shall see that the KW duality sym-metry property,Eq.(2.10),results in the beta-function being covariant under the operation g→d(g):β(d(g))=d′(g)β(g)(2.16)To prove it let us evaluateβ(d(g)).Then Eq.(2.14)yieldsβ(d(g))=2s(˜g)(1−s(˜g))s(g)(1+s(g))(ds(˜g)/d˜g)(2.18)The derivative in the r.h.s.of Eq.(2.18)should be rewritten in terms of s(g)and d(g).It may be easily done by applying Eq.(2.15):ds(˜g)d˜g 1s2(g)1g +24g1g−12(2.20)Combining this Pad´e-approximant with the definition of d(g),Eq.(2.15),one is led tod(g)=43g−35the appropriate expansion parameter.Having been equipped with these formulas,one may easily calculate the beta-functionβ(g)as a power series in g−1.Inserting Eq.(A.10)into Eq.(2.14)and performing simple but somewhat cumbersome calculations,we are led to the desired asymptotic expansion forβ(g)β(g)=−2g+32−64/g+512/g2+512/g3−30720/g4−172032/g5+32768/g6−172032/g7+32768/g8+0(g−9)(3.1)¿From Eq.(3.1)it follows that in the large-g limitβ(g)→−2g+32,whilst in the weak coupling regime one has for g→0:β(g)→+2g[2–4,23,24].It implies that the continuous functionβ(g)changes sign at least once at somefixed point g∗.Let us get some numerical estimates for g∗+now,from Eq.(3.1),and compare these results with those found from the HT series expansions and those of the four-loop RG calculations. In the standard perturbative approach to quantumfield theory atfixed dimension one must apply some resummation technique to the expansions ofβ(g)and other RG-functions.It is interesting that at least in low orders of perturbation theory the1/g-expansion,Eq.(3.1), does not require the application of a resummation technique.The most reliable numerical estimates of g∗+were obtained by means of the straightforward solution of the equation β(g∗)=0,from Eq.(3.1),taken within the g−6-approximation(without the last two terms in g−7and g−8).Thefive(and best)subsequent approximation are as followsg∗(1) +=16;g∗(2)+=13.6568;g∗(3)+=15.0044;g∗(4) +=15.0784;g∗(5)+=14.7632(3.2)Here the index k in g∗(k)+indicates that k+1terms are retained in thefixed point equationunder discussion.These estimates exhibit a regular behavior,the last value being in very good agreement with the most recent estimate g∗+=14.700±0.017obtained for the square lattice[14,15].The estimates obtained after taking into account the g−7and g−8terms differ significantly from the above values.This is apparently an indication that1/g-series also require the application of some resummation technique.Another approach to obtain a numerical estimate for g∗+is a straightforward solution of Eq.(A.10),given in the Appendix,after setting s=1.In contrast to thefixed point equation,β(g)=0,it yields a rather poor value of the renormalized coupling constant, g∗+=12.533,compared to the value reported in[3,14,15].It is interesting to compare our results with those obtained from the beta-function of the 2D Ising model and computed in the four-loop approximation,known to provide more or less satisfactory results for the critical indices[2–4]:β(v)=2v−2v2+1.432346v3−1.861533v4+3.1647764v5+0(v6)(3.3) To obtain the beta-function in our normalization we have to change variables[24]g=8π3β(v)(3.4)The analysis based on the Pad´e-Borel method of resummation of asymptotic series yields g∗+=15.08±2.5[24],which slightly exceeds the best values obtained from the HT seriescalculations:g ∗+=14.70±0.017[14,15];g ∗+=14.67±0.04[24](it is tempting to conjecturethat g ∗+=14π3[23].IV.ISOTHERMAL COUPLING CONSTANT AND CRITICAL AMPLITUDES The two preceeding Sections were devoted to computing the approximate value of therenormalized coupling constant g ∗+at h =0in the isochore limit.Here we remark thatin two dimensions there is a possibility of calculating the exact value of the renormalizedcoupling constant g ∗c in the isothermal limit,i.e.at the Curie point in an applied magneticfield,namelyg ∗c =lim h →0g (T =T c ,h )(4.1)by virtue of Cardy’formula [26].It is in fact essential to stress that,in contrast to otherisothermal critical amplitudes,g ∗c is fixed by this formula,which reads [26]c =3πh 2(2−η4−η;χ(h )=C c h 2η−4(4−η)2C c (f c 1)2(4.5)On the other hand,from Eq.s (2.6)and (4.3)it is seen that the correlation length ξdropsout of the product g ∗c c :cg ∗c =3π(4−η)2h 2 −∂2χ/∂h 2χ2(4.6)Inserting Eq.(4.4)into the r.h.s.of Eq.(4.6),one obtains the renormalized coupling constant value at the end point of the isothermal lineg∗c=6π3π(4−η)2(4.8)¿From Eq.(4.8)it follows that what we actually found,by virtue of Cardy’s formula,is only the product C c(f c1)2.To compute these quantities separately one needs more powerful techniques.In some seminal papers[27–29]it was shown how to compute the isothermal amplitudes by making use of the Thermodynamic Bethe Ansatz and within the framework of the form-factor approach.In particular,in his paper Fateev obtained the following remarkable result [27]f c1=Γ(2/3)Γ(8/15)4π2Γ(3/4)Γ2(13/16)]4/15=0.2270194675(4.9)¿From Eq.s(4.8)and(4.9)it follows thatC c=0.0731998414(4.10) All these results allow one to compute exactly the two following universal combinations[13] (see also[23]and[30])Q1=C cδC c ( f c1The exact values found provide a good opportunity to test the numerical results ob-tained from the HT series expansions and from Monte Carlo simulations.The fair esti-mates obtained from the analysis of HT series in the2D Ising model on the square lattice yield f c1=0.233and C c=0.0706([13]),whilst the exact results are given by Eq.s(4.9) and(4.10).As for the universal combinations Q1,2,the series expansion analysis yields Q1=0.88023,Q2=2.88[13].Notice in conclusion that Eq.s(4.7)and(4.8)hold good also for the general case of the 2D gΦ40(N)-symmetric model for−2<N<2,in particular for the minimal models of con-formalfield theory corresponding to the discrete values of N:N=2cos(πall the corrections to the scaling laws in the2D Ising model are analytical.For instance, the susceptibility near T c is given byχ=C+τ−7/4+C+1τ−3/4+...[31].On the other hand, corrections to scaling are known to be powers ofτων[23].All this would lead toω=1, in obvious contradiction to conformalfield theory.Moreover,the spectrum of conformal dimensions of the2D Ising model consists of just three numbers,these being(0,1)are as follows[15]Tχ=1+4K+12K2+104K3/3+92K4+3608K5/15+3056K6/5+484528K7/315+400012K8/105(A.1)µ2=4K+32K2+488K3/3+2048K4/3+38168K5/15+394624K6/45+8994736K7/315+28064768K8/315(A.2)χ′′hh=−2−32K−264K2−4864K3/3−8232K4−553024K5/15−2259616K6/15−180969728K7/315−217858792K8/105(A.3) withχ′′hh being the second derivative of the homogenous susceptibility with respect to a magneticfield h;χandµ2were defined in Sect.II.The standard RG equation for the effective temperature T is given byξdT2dχ(A.5)The key observation for computingξis to make use of the new variable s= exp(2K)tanh(K).One has to substitute Eq.s(A.1)and(A.2)into Eq.(A.5)and then to rewrite the expression obtained in terms of s.At this order of approximation the proce-dure gives the resultξ2=s+2s2+3s3+4s4+5s5+6s6+7s7+8s8+0(s9)=sτ;τ=T−T cln(√4ln(√d lnξ=βIsing(s)=2s(1−s)1d lnξ=βIsing(T)=T21−2exp(−2/T)−2exp(−4/T)+2exp(−6/T)+exp(−8/T)ds|s=1=1(A.12)(iii)the beta-function is covariant under the duality transformation,namelyβIsing(s)=−s2βIsing(12sinh2T(A.14)does not satisfy either properties(ii)and(iv)[35].To end with,we prove some useful relations concerning the duality transformation˜g= d(g)introduced in Sect.II.i)Let us show that d′(g∗)=±1.First of all,from the definitiond(g)≡s−1(1d(d(g))=g;d(g∗)=g∗d′(d(g))d′(g)=1;g=g∗d′(g∗)d′(g∗)=1(A.16) showing that indeed d′(g∗)=±1.iii)Differentiating Eq.(A.16)(second from top)with respect to g one obtainsd′[d(g)]d′(g)=1;d′′[d(g)]d′(g)2+d′[d(g)]d′′(g)=0(A.17) and at thefixed point we arrive at(d(g∗)=g∗):d′′(g∗)(d(g∗))2+d′(g∗)d′′(g∗)=0(A.18) If d′(g∗)=−1we have an identity d′′(g∗)=d′′(g∗).In the opposite case from d′(g∗)=+1 it follows that d′′(g∗)=0.Proceeding in the same way,it is easy to see that all higher derivatives vanish identicatically at thefixed point.This,alone,does not imply that d(g)≡g,since it could be that d(g)=g+f(g)where f(g)is some nonanalytic function having vanishing derivatives at g∗.Let us then assume that f(g)is thefirst term of an asymptotic expansion of d(g)around thefixed point,so that f(g)→0as g→g∗.Then remembering that d(d(g))=g,g=d(d(g))=d(g+f(g))=g+f(g)+f(g+f(g))≃g+2f(g)(A.19) and we do obtain f(g)≡0.REFERENCES∗(permanent address)∗∗On leave of absence from:A.F.Ioffe Physical&Technical Institute,Russian Academy of Sciences,Polytechnicheskaya str.26,194021St.Petersburg(Russia)(permanent ad-dress)[1]McCoy B.M.,HEPTH/9403084[2]Nickel B.G.,Meiron D.I.and Baker G.A.Jr.,Compilation of2-pt and4-pt graphs forcontinuous spin models,University of Guelph report(1977)[3]Baker G.A.Jr.,Nickel B.G.and Meiron D.I.,Phys.Rev.B17,1365(1978)[4]Le Guillou J.C.and Zinn-Justin J.,Phys.Rev.B21,3976(1980)[5]Kramers H.A.and Wannier G.H.,Phys.Rev.60,252(1941)[6]Savit R.,Rev.Mod.Phys.52,453(1980)[7]Kogut J.B.,Rev.Mod.Phys.51,659(1979)[8]Shalaev B.N.,Phys.Rep.237,129(1994)[9]Dotsenko V.S.,Sov.Phys-JETP75,1083(1978)[10]Dolan L.and Grady M.,Phys.Rev.D25,1587(1982)[11]Damgaard P.H.and Haagensen P.E.,J.Phys.A:Math.Gen.30,4681(1997)[12]Correia J.D.,hep-th/9806229[13]Tarko H.B.and Fisher M.E.,Phys.Rev.B11,1217(1975)[14]Shun-Yong Zinn,Sheng-Nan Lai and Fisher M.E.,Phys.Rev.E54,1176(1996)[15]Butera P.and Comi M.,Phys.Rev.B54,15828(1996)[16]Butera P.and Comi M.,Phys.Rev.E55,6391(1997)[17]Reisz T.,Phys.Lett.B360,77(1995)[18]Guida R.and Zinn-Justin J.,Nucl.Phys.B489,626(1997)[19]Sokolov A.I.,Phys Solid State40,N7(1998)[20]Sokolov A.I.,Orlov E.V.and Ul’kov V.A.,Phys.Lett.A227,255(1997)[21]Shun-Yong Zinn and Fisher M.E.,Physica A226,168(1996)[22]Kleinert H.,Phys.Rev.D57,2264(1998);cond-mat/9801167[23]Zinn-Justin J.,Quantum Field Theory and Critical Phenomena,3rd ed.,(ClarendonPress,Oxford)(1996)[24]Baker G.A.Jr.,Phys.Rev.B15,1552(1977)[25]Wegner F.,Phys.Rev.B5,4529(1972);Phys.Rev.B6,1891(1972)[26]Cardy J.L.,Phys.Rev.Lett.26,2709(1988)[27]Fateev V.A.,Phys.Lett.B324,45(1994)[28]Mussardo G.in Integrable Quantum Field Theories NATO ASI Series B:Physics310173(1993)(Plenum Press,New York)(1993)[29]Delfino G.and G.Mussardo,Nucl.Phys.B455,724(1995)[30]Privman V.,P.C.Hohenberg and A.Aharony,Phase Transitions and Critical Phenom-ena,Vol.14,p.4,C.Domb and J.Lebowitz eds.,(Academic Press,New York)(1991)[31]Barouch E.,B.M.McCoy and T.T.Wu,Phys.Rev.Lett.31,1409(1973)[32]Pelissetto A.and E.Vicari Nucl.Phys.B540[FS],639(1999)[33]Wegner F.J.,Nucl.Phys.B316,663(1989)[34]Baxter R.J.,Exactly Solved Models in Statistical Mechanics(Academic Press,New York)(1982),p.120[35]Migdal A.A.,Sov.Phys.JETP42,413(1975)。

计量经济学中英文词汇对照

计量经济学中英文词汇对照

Controlled experiments Conventional depth Convolution Corrected factor Corrected mean Correction coefficient Correctness Correlation coefficient Correlation index Correspondence Counting Counts Covaห้องสมุดไป่ตู้iance Covariant Cox Regression Criteria for fitting Criteria of least squares Critical ratio Critical region Critical value
Asymmetric distribution Asymptotic bias Asymptotic efficiency Asymptotic variance Attributable risk Attribute data Attribution Autocorrelation Autocorrelation of residuals Average Average confidence interval length Average growth rate BBB Bar chart Bar graph Base period Bayes' theorem Bell-shaped curve Bernoulli distribution Best-trim estimator Bias Binary logistic regression Binomial distribution Bisquare Bivariate Correlate Bivariate normal distribution Bivariate normal population Biweight interval Biweight M-estimator Block BMDP(Biomedical computer programs) Boxplots Breakdown bound CCC Canonical correlation Caption Case-control study Categorical variable Catenary Cauchy distribution Cause-and-effect relationship Cell Censoring

数学专业词汇及翻译

数学专业词汇及翻译

一、字母顺序表 (1)二、常用的数学英语表述 (7)三、代数英语(高端) (13)一、字母顺序表1、数学专业词汇Aabsolute value 绝对值 accept 接受 acceptable region 接受域additivity 可加性 adjusted 调整的 alternative hypothesis 对立假设analysis 分析 analysis of covariance 协方差分析 analysis of variance 方差分析 arithmetic mean 算术平均值 association 相关性 assumption 假设 assumption checking 假设检验availability 有效度average 均值Bbalanced 平衡的 band 带宽 bar chart 条形图beta-distribution 贝塔分布 between groups 组间的 bias 偏倚 binomial distribution 二项分布 binomial test 二项检验Ccalculate 计算 case 个案 category 类别 center of gravity 重心 central tendency 中心趋势 chi-square distribution 卡方分布 chi-square test 卡方检验 classify 分类cluster analysis 聚类分析 coefficient 系数 coefficient of correlation 相关系数collinearity 共线性 column 列 compare 比较 comparison 对照 components 构成,分量compound 复合的 confidence interval 置信区间 consistency 一致性 constant 常数continuous variable 连续变量 control charts 控制图 correlation 相关 covariance 协方差 covariance matrix 协方差矩阵 critical point 临界点critical value 临界值crosstab 列联表cubic 三次的,立方的 cubic term 三次项 cumulative distribution function 累加分布函数 curve estimation 曲线估计Ddata 数据default 默认的definition 定义deleted residual 剔除残差density function 密度函数dependent variable 因变量description 描述design of experiment 试验设计 deviations 差异 df.(degree of freedom) 自由度 diagnostic 诊断dimension 维discrete variable 离散变量discriminant function 判别函数discriminatory analysis 判别分析distance 距离distribution 分布D-optimal design D-优化设计Eeaqual 相等 effects of interaction 交互效应 efficiency 有效性eigenvalue 特征值equal size 等含量equation 方程error 误差estimate 估计estimation of parameters 参数估计estimations 估计量evaluate 衡量exact value 精确值expectation 期望expected value 期望值exponential 指数的exponential distributon 指数分布 extreme value 极值F factor 因素,因子 factor analysis 因子分析 factor score 因子得分 factorial designs 析因设计factorial experiment 析因试验fit 拟合fitted line 拟合线fitted value 拟合值 fixed model 固定模型 fixed variable 固定变量 fractional factorial design 部分析因设计 frequency 频数 F-test F检验 full factorial design 完全析因设计function 函数Ggamma distribution 伽玛分布 geometric mean 几何均值 group 组Hharmomic mean 调和均值 heterogeneity 不齐性histogram 直方图 homogeneity 齐性homogeneity of variance 方差齐性 hypothesis 假设 hypothesis test 假设检验Iindependence 独立 independent variable 自变量independent-samples 独立样本 index 指数 index of correlation 相关指数 interaction 交互作用 interclass correlation 组内相关 interval estimate 区间估计 intraclass correlation 组间相关 inverse 倒数的iterate 迭代Kkernal 核 Kolmogorov-Smirnov test柯尔莫哥洛夫-斯米诺夫检验 kurtosis 峰度Llarge sample problem 大样本问题 layer 层least-significant difference 最小显著差数 least-square estimation 最小二乘估计 least-square method 最小二乘法 level 水平 level of significance 显著性水平 leverage value 中心化杠杆值 life 寿命 life test 寿命试验 likelihood function 似然函数 likelihood ratio test 似然比检验linear 线性的 linear estimator 线性估计linear model 线性模型 linear regression 线性回归linear relation 线性关系linear term 线性项logarithmic 对数的logarithms 对数 logistic 逻辑的 lost function 损失函数Mmain effect 主效应 matrix 矩阵 maximum 最大值 maximum likelihood estimation 极大似然估计 mean squared deviation(MSD) 均方差 mean sum of square 均方和 measure 衡量 media 中位数 M-estimator M估计minimum 最小值 missing values 缺失值 mixed model 混合模型 mode 众数model 模型Monte Carle method 蒙特卡罗法 moving average 移动平均值multicollinearity 多元共线性multiple comparison 多重比较 multiple correlation 多重相关multiple correlation coefficient 复相关系数multiple correlation coefficient 多元相关系数 multiple regression analysis 多元回归分析multiple regression equation 多元回归方程 multiple response 多响应 multivariate analysis 多元分析Nnegative relationship 负相关 nonadditively 不可加性 nonlinear 非线性 nonlinear regression 非线性回归 noparametric tests 非参数检验 normal distribution 正态分布null hypothesis 零假设 number of cases 个案数Oone-sample 单样本 one-tailed test 单侧检验 one-way ANOVA 单向方差分析 one-way classification 单向分类 optimal 优化的optimum allocation 最优配制 order 排序order statistics 次序统计量 origin 原点orthogonal 正交的 outliers 异常值Ppaired observations 成对观测数据paired-sample 成对样本parameter 参数parameter estimation 参数估计 partial correlation 偏相关partial correlation coefficient 偏相关系数 partial regression coefficient 偏回归系数 percent 百分数percentiles 百分位数 pie chart 饼图 point estimate 点估计 poisson distribution 泊松分布polynomial curve 多项式曲线polynomial regression 多项式回归polynomials 多项式positive relationship 正相关 power 幂P-P plot P-P概率图predict 预测predicted value 预测值prediction intervals 预测区间principal component analysis 主成分分析 proability 概率 probability density function 概率密度函数 probit analysis 概率分析 proportion 比例Qqadratic 二次的 Q-Q plot Q-Q概率图 quadratic term 二次项 quality control 质量控制 quantitative 数量的,度量的 quartiles 四分位数Rrandom 随机的 random number 随机数 random number 随机数 random sampling 随机取样random seed 随机数种子 random variable 随机变量 randomization 随机化 range 极差rank 秩 rank correlation 秩相关 rank statistic 秩统计量 regression analysis 回归分析regression coefficient 回归系数regression line 回归线reject 拒绝rejection region 拒绝域 relationship 关系 reliability 可*性 repeated 重复的report 报告,报表 residual 残差 residual sum of squares 剩余平方和 response 响应risk function 风险函数 robustness 稳健性 root mean square 标准差 row 行 run 游程run test 游程检验Sample 样本 sample size 样本容量 sample space 样本空间 sampling 取样 sampling inspection 抽样检验 scatter chart 散点图 S-curve S形曲线 separately 单独地 sets 集合sign test 符号检验significance 显著性significance level 显著性水平significance testing 显著性检验 significant 显著的,有效的 significant digits 有效数字 skewed distribution 偏态分布 skewness 偏度 small sample problem 小样本问题 smooth 平滑 sort 排序 soruces of variation 方差来源 space 空间 spread 扩展square 平方 standard deviation 标准离差 standard error of mean 均值的标准误差standardization 标准化 standardize 标准化 statistic 统计量 statistical quality control 统计质量控制 std. residual 标准残差 stepwise regression analysis 逐步回归 stimulus 刺激 strong assumption 强假设 stud. deleted residual 学生化剔除残差stud. residual 学生化残差 subsamples 次级样本 sufficient statistic 充分统计量sum 和 sum of squares 平方和 summary 概括,综述Ttable 表t-distribution t分布test 检验test criterion 检验判据test for linearity 线性检验 test of goodness of fit 拟合优度检验 test of homogeneity 齐性检验 test of independence 独立性检验 test rules 检验法则 test statistics 检验统计量 testing function 检验函数 time series 时间序列 tolerance limits 容许限total 总共,和 transformation 转换 treatment 处理 trimmed mean 截尾均值 true value 真值 t-test t检验 two-tailed test 双侧检验Uunbalanced 不平衡的 unbiased estimation 无偏估计 unbiasedness 无偏性 uniform distribution 均匀分布Vvalue of estimator 估计值 variable 变量 variance 方差 variance components 方差分量 variance ratio 方差比 various 不同的 vector 向量Wweight 加权,权重 weighted average 加权平均值 within groups 组内的ZZ score Z分数2. 最优化方法词汇英汉对照表Aactive constraint 活动约束 active set method 活动集法 analytic gradient 解析梯度approximate 近似 arbitrary 强制性的 argument 变量 attainment factor 达到因子Bbandwidth 带宽 be equivalent to 等价于 best-fit 最佳拟合 bound 边界Ccoefficient 系数 complex-value 复数值 component 分量 constant 常数 constrained 有约束的constraint 约束constraint function 约束函数continuous 连续的converge 收敛 cubic polynomial interpolation method三次多项式插值法 curve-fitting 曲线拟合Ddata-fitting 数据拟合 default 默认的,默认的 define 定义 diagonal 对角的 direct search method 直接搜索法 direction of search 搜索方向 discontinuous 不连续Eeigenvalue 特征值 empty matrix 空矩阵 equality 等式 exceeded 溢出的Ffeasible 可行的 feasible solution 可行解 finite-difference 有限差分 first-order 一阶GGauss-Newton method 高斯-牛顿法 goal attainment problem 目标达到问题 gradient 梯度 gradient method 梯度法Hhandle 句柄 Hessian matrix 海色矩阵Independent variables 独立变量inequality 不等式infeasibility 不可行性infeasible 不可行的initial feasible solution 初始可行解initialize 初始化inverse 逆 invoke 激活 iteration 迭代 iteration 迭代JJacobian 雅可比矩阵LLagrange multiplier 拉格朗日乘子 large-scale 大型的 least square 最小二乘 least squares sense 最小二乘意义上的 Levenberg-Marquardt method 列文伯格-马夸尔特法line search 一维搜索 linear 线性的 linear equality constraints 线性等式约束linear programming problem 线性规划问题 local solution 局部解M medium-scale 中型的 minimize 最小化 mixed quadratic and cubic polynomialinterpolation and extrapolation method 混合二次、三次多项式内插、外插法multiobjective 多目标的Nnonlinear 非线性的 norm 范数Oobjective function 目标函数 observed data 测量数据 optimization routine 优化过程optimize 优化 optimizer 求解器 over-determined system 超定系统Pparameter 参数 partial derivatives 偏导数 polynomial interpolation method 多项式插值法Qquadratic 二次的 quadratic interpolation method 二次内插法 quadratic programming 二次规划Rreal-value 实数值 residuals 残差 robust 稳健的 robustness 稳健性,鲁棒性S scalar 标量 semi-infinitely problem 半无限问题 Sequential Quadratic Programming method 序列二次规划法 simplex search method 单纯形法 solution 解 sparse matrix 稀疏矩阵 sparsity pattern 稀疏模式 sparsity structure 稀疏结构 starting point 初始点 step length 步长 subspace trust region method 子空间置信域法 sum-of-squares 平方和 symmetric matrix 对称矩阵Ttermination message 终止信息 termination tolerance 终止容限 the exit condition 退出条件 the method of steepest descent 最速下降法 transpose 转置Uunconstrained 无约束的 under-determined system 负定系统Vvariable 变量 vector 矢量Wweighting matrix 加权矩阵3 样条词汇英汉对照表Aapproximation 逼近 array 数组 a spline in b-form/b-spline b样条 a spline of polynomial piece /ppform spline 分段多项式样条Bbivariate spline function 二元样条函数 break/breaks 断点Ccoefficient/coefficients 系数cubic interpolation 三次插值/三次内插cubic polynomial 三次多项式 cubic smoothing spline 三次平滑样条 cubic spline 三次样条cubic spline interpolation 三次样条插值/三次样条内插 curve 曲线Ddegree of freedom 自由度 dimension 维数Eend conditions 约束条件 input argument 输入参数 interpolation 插值/内插 interval取值区间Kknot/knots 节点Lleast-squares approximation 最小二乘拟合Mmultiplicity 重次 multivariate function 多元函数Ooptional argument 可选参数 order 阶次 output argument 输出参数P point/points 数据点Rrational spline 有理样条 rounding error 舍入误差(相对误差)Sscalar 标量 sequence 数列(数组) spline 样条 spline approximation 样条逼近/样条拟合spline function 样条函数 spline curve 样条曲线 spline interpolation 样条插值/样条内插 spline surface 样条曲面 smoothing spline 平滑样条Ttolerance 允许精度Uunivariate function 一元函数Vvector 向量Wweight/weights 权重4 偏微分方程数值解词汇英汉对照表Aabsolute error 绝对误差 absolute tolerance 绝对容限 adaptive mesh 适应性网格Bboundary condition 边界条件Ccontour plot 等值线图 converge 收敛 coordinate 坐标系Ddecomposed 分解的 decomposed geometry matrix 分解几何矩阵 diagonal matrix 对角矩阵 Dirichlet boundary conditions Dirichlet边界条件Eeigenvalue 特征值 elliptic 椭圆形的 error estimate 误差估计 exact solution 精确解Ggeneralized Neumann boundary condition 推广的Neumann边界条件 geometry 几何形状geometry description matrix 几何描述矩阵 geometry matrix 几何矩阵 graphical user interface(GUI)图形用户界面Hhyperbolic 双曲线的Iinitial mesh 初始网格Jjiggle 微调LLagrange multipliers 拉格朗日乘子Laplace equation 拉普拉斯方程linear interpolation 线性插值 loop 循环Mmachine precision 机器精度 mixed boundary condition 混合边界条件NNeuman boundary condition Neuman边界条件 node point 节点 nonlinear solver 非线性求解器 normal vector 法向量PParabolic 抛物线型的 partial differential equation 偏微分方程 plane strain 平面应变 plane stress 平面应力 Poisson's equation 泊松方程 polygon 多边形 positive definite 正定Qquality 质量Rrefined triangular mesh 加密的三角形网格 relative tolerance 相对容限 relative tolerance 相对容限 residual 残差 residual norm 残差范数Ssingular 奇异的二、常用的数学英语表述1.Logic∃there exist∀for allp⇒q p implies q / if p, then qp⇔q p if and only if q /p is equivalent to q / p and q are equivalent2.Setsx∈A x belongs to A / x is an element (or a member) of Ax∉A x does not belong to A / x is not an element (or a member) of AA⊂B A is contained in B / A is a subset of BA⊃B A contains B / B is a subset of AA∩B A cap B / A meet B / A intersection BA∪B A cup B / A join B / A union BA\B A minus B / the diference between A and BA×B A cross B / the cartesian product of A and B3. Real numbersx+1 x plus onex-1 x minus onex±1 x plus or minus onexy xy / x multiplied by y(x - y)(x + y) x minus y, x plus yx y x over y= the equals signx = 5 x equals 5 / x is equal to 5x≠5x (is) not equal to 5x≡y x is equivalent to (or identical with) yx ≡ y x is not equivalent to (or identical with) yx > y x is greater than yx≥y x is greater than or equal to yx < y x is less than yx≤y x is less than or equal to y0 < x < 1 zero is less than x is less than 10≤x≤1zero is less than or equal to x is less than or equal to 1| x | mod x / modulus xx 2 x squared / x (raised) to the power 2x 3 x cubedx 4 x to the fourth / x to the power fourx n x to the nth / x to the power nx −n x to the (power) minus nx (square) root x / the square root of xx 3 cube root (of) xx 4 fourth root (of) xx n nth root (of) x( x+y ) 2 x plus y all squared( x y ) 2 x over y all squaredn! n factorialx ^ x hatx ¯ x barx ˜x tildex i xi / x subscript i / x suffix i / x sub i∑ i=1 n a i the sum from i equals one to n a i / the sum as i runs from 1 to n of the a i4. Linear algebra‖ x ‖the norm (or modulus) of xOA →OA / vector OAOA ¯ OA / the length of the segment OAA T A transpose / the transpose of AA −1 A inverse / the inverse of A5. Functionsf( x ) fx / f of x / the function f of xf:S→T a function f from S to Tx→y x maps to y / x is sent (or mapped) to yf'( x ) f prime x / f dash x / the (first) derivative of f with respect to xf''( x ) f double-prime x / f double-dash x / the second derivative of f with r espect to xf'''( x ) triple-prime x / f triple-dash x / the third derivative of f with respect to xf (4) ( x ) f four x / the fourth derivative of f with respect to x∂f ∂ x 1the partial (derivative) of f with respect to x1∂ 2 f ∂ x 1 2the second partial (derivative) of f with respect to x1∫ 0 ∞the integral from zero to infinitylim⁡x→0 the limit as x approaches zerolim⁡x→0 + the limit as x approaches zero from abovelim⁡x→0 −the limit as x approaches zero from belowlog e y log y to the base e / log to the base e of y / natural log (of) yln⁡y log y to the base e / log to the base e of y / natural log (of) y一般词汇数学mathematics, maths(BrE), math(AmE)公理axiom定理theorem计算calculation运算operation证明prove假设hypothesis, hypotheses(pl.)命题proposition算术arithmetic加plus(prep.), add(v.), addition(n.)被加数augend, summand加数addend和sum减minus(prep.), subtract(v.), subtraction(n.)被减数minuend减数subtrahend差remainder乘times(prep.), multiply(v.), multiplication(n.)被乘数multiplicand, faciend乘数multiplicator积product除divided by(prep.), divide(v.), division(n.)被除数dividend除数divisor商quotient等于equals, is equal to, is equivalent to 大于is greater than小于is lesser than大于等于is equal or greater than小于等于is equal or lesser than运算符operator数字digit数number自然数natural number整数integer小数decimal小数点decimal point分数fraction分子numerator分母denominator比ratio正positive负negative零null, zero, nought, nil十进制decimal system二进制binary system十六进制hexadecimal system权weight, significance进位carry截尾truncation四舍五入round下舍入round down上舍入round up有效数字significant digit无效数字insignificant digit代数algebra公式formula, formulae(pl.)单项式monomial多项式polynomial, multinomial系数coefficient未知数unknown, x-factor, y-factor, z-factor 等式,方程式equation一次方程simple equation二次方程quadratic equation三次方程cubic equation四次方程quartic equation不等式inequation阶乘factorial对数logarithm指数,幂exponent乘方power二次方,平方square三次方,立方cube四次方the power of four, the fourth power n次方the power of n, the nth power开方evolution, extraction二次方根,平方根square root三次方根,立方根cube root四次方根the root of four, the fourth root n次方根the root of n, the nth root集合aggregate元素element空集void子集subset交集intersection并集union补集complement映射mapping函数function定义域domain, field of definition值域range常量constant变量variable单调性monotonicity奇偶性parity周期性periodicity图象image数列,级数series微积分calculus微分differential导数derivative极限limit无穷大infinite(a.) infinity(n.)无穷小infinitesimal积分integral定积分definite integral不定积分indefinite integral有理数rational number无理数irrational number实数real number虚数imaginary number复数complex number矩阵matrix行列式determinant几何geometry点point线line面plane体solid线段segment射线radial平行parallel相交intersect角angle角度degree弧度radian锐角acute angle直角right angle钝角obtuse angle平角straight angle周角perigon底base边side高height三角形triangle锐角三角形acute triangle直角三角形right triangle直角边leg斜边hypotenuse勾股定理Pythagorean theorem钝角三角形obtuse triangle不等边三角形scalene triangle等腰三角形isosceles triangle等边三角形equilateral triangle四边形quadrilateral平行四边形parallelogram矩形rectangle长length宽width附:在一个分数里,分子或分母或两者均含有分数。

Uncertainty measures on probability intervals from the imprecise Dirichlet model

Uncertainty measures on probability intervals from the imprecise Dirichlet model

Uncertainty measures on probability intervals from theimprecise Dirichlet modelJ.ABELLA´N*Department of Computer Science and Artificial Intelligence,University of Granada,Granada 18071,Spain(Received 6February 2006;in final form 3March 2006)When we use a mathematical model to represent information,we can obtain a closed and convex set of probability distributions,also called a credal set.This type of representation involves two types of uncertainty called conflict (or randomness )and non-specificity ,respectively.The imprecise Dirichlet model (IDM)allows us to carry out inference about the probability distribution of a categorical variable obtaining a set of a special type of credal set (probability intervals).In this paper,we shall present tools for obtaining the uncertainty functions on probability intervals obtained with the IDM,which can enable these functions in any application of this model to be calculated.Keywords :Imprecise probabilities;Credal sets;Uncertainty;Entropy;Conflict;Imprecise Dirichlet model1.IntroductionSince the amount of information obtained by any action is measured by a reduction in uncertainty,the concept of uncertainty is intricately connected to the concept of information.The concept of ‘information-based uncertainty’(Klir and Wierman 1998)is related to information deficiencies such as the information being incomplete,imprecise,fragmentary,not fully reliable,vague,contradictory or deficient,and this may result in different types of uncertainty.This paper is solely concerned with the information conceived in terms of uncertainty reduction,unlike the term ‘information’as it is used in the theory of computability or in terms of logic.In classic information theory,Shannon’s entropy (1948)is the tool used to quantify uncertainty.This function has certain desirable properties and has been used as the starting point when looking for another function to measure the amount of uncertainty in situations in which a probabilistic representation is not suitable.Many mathematical imprecise probability theories for representing information-based uncertainty are based on a generalization of the probability theory:e.g.Dempster–Shafer’s theory (DST)(Dempster 1967,Shafer 1976),interval-valued probabilities (de Campos et al.1994),order-2capacities (Choquet 1953/1954),upper–lower probabilities (Suppes 1974,Fine 1983,International Journal of General SystemsISSN 0308-1079print/ISSN 1563-5104online q 2006Taylor &Francis/journalsDOI:10.1080/03081070600687643*Email:jabellan@decsai.ugr.esInternational Journal of General Systems ,Vol.35,No.5,October 2006,509–5281988)or general convex sets of probability distributions (Good 1962,Levi 1980,Walley 1991,Berger 1994).Each of these represents a type of credal set that is a closed and convex set of probability distributions with a finite set of extreme points.In the DST,Yager (1983)distinguishes between two types of uncertainty:one is associated with cases where the information focuses on sets with empty intersections,and the other is associated with cases where the information focuses on sets where the cardinality is greater than one.These are called conflict and non-specificity ,respectively.The study of uncertainty measures in the DST is the starting point for the study of these measures on more general theories.In any of these theories,it is justifiable that a measure capable of measuring the uncertainty represented by a credal set must quantify the parts of conflict and non-specificity.More recently,Abella´n and Moral (2005b)and Klir and Smith (2001)justified the use of maximum entropy on credal sets as a good measure of total uncertainty.The problem lies in separating these functions into others that really do measure the conflict and non-specificityparts by using a credal set to represent the information.Abella´n et al.(2006)managed to split maximum entropy into functions that are capable of coherently measuring the conflict and non-specificity of a credal set P ;and also as algorithms in order to facilitate their calculationin order-2capacities (Abella´n and Moral 2005a,2006)so that S *ðP Þ¼S *ðP ÞþðS *2S *ÞðP Þ;where S *represents maximum entropy and S *represents minimum entropy on a credal set P ;with S *ðP Þcoherently quantifying the conflict part of a credal set and ðS *2S *ÞðP Þthe non-specificity part of a credal set.A natural way of representing knowledge is with probability intervals (Campos et al.1994).In this paper,we shall work with a special type of probability intervals obtained using the imprecise Dirichlet model (IDM).The main use of IDM is to infer about acategorical variable.Abella´n and Moral (2003b,2005b)recently used IDM to join uncertainty measures in classification (an important problem in the field of machine learning ).In this paper,we shall study IDM probability intervals and we shall prove that,while they can be represented by belief functions,they are not the only type of credal set belonging to belief functions and probability intervals.In addition,we shall present an algorithm that obtains the maximum entropy for this type of interval;we shall demonstrate a property that will enable us rapidly to obtain the minimum entropy for this type of interval;and using the fact that they represent a special type of belief function,we shall directly obtain the value of the Hartley measure on them.In Section 2of this paper,we shall introduce the most important imprecise probability theories and distinguish between probability intervals and belief functions.In Section 3,we shall present the IDM and its main properties and shall also examine the situation of IDM probability intervals in relation to other imprecise probability theories.In Section 4,we shall explore uncertainty measures on credal sets.In Section 5,we shall outline some procedures and algorithms for obtaining the values of the main uncertainty measures on IDM probability intervals and practical examples.Conclusions are presented in Section 6.J.Abella´n 5102.Theories of imprecise probabilities2.1Credal setsAll theories of imprecise probabilities that are based on classical set theory share some common characteristics(see Walley1991,Klir2006).One of them is that evidence within each theory is fully described by a lower probability function P*on afinite set X or, alternatively,by an upper probability function P*on X.These functions are always regular monotone measures(Wang and Klir1992)that are superadditive and subadditive, respectively,andX x[X P*ð{x}Þ<1;Xx[XP*ð{x}Þ>1:ð1ÞIn the various special theories of uncertainty,they possess additional special properties. When evidence is expressed(at the most general level)in terms of an arbitrary credal set,P of probability distribution functions p,on afinite set X(Kyburg1987),functions P*and P* associated with P are determined for each set A#X by the formulaeP*ðAÞ¼infp[P Xx[Apð{x}Þ;P*ðAÞ¼supp[PXx[Apð{x}Þ:ð2ÞSince for each p[P and each A#X,it follows thatP*ðAÞ¼12P*ðX2AÞ:ð3ÞOwing to this property,functions P*and P*are called dual(or conjugate).One of them is sufficient for capturing given evidence;the other one is uniquely determined by equation(3). It is common to use the lower probability function to capture the evidence.As is well known (Chateauneuf and Jaffray1989,Grabisch2000)any given lower probability function P*is uniquely represented by a set-valued function m for which mðYÞ¼0andXA[‘ðXÞmðAÞ¼1;ð4Þwhere we note‘(X)as the power set of X.Any set A#X for which mðAÞ–0is often called a focal element,and the set of all focal elements with the values assigned to them by function m is called a body of evidence.Function m is called a Mo¨bius representation of P*when it is obtained for all A#X via the Mo¨bius transformmðAÞ¼XB j B#Að21Þj A2B j P*ðBÞ:ð5ÞThe inverse transform is defined for all A#X by the formulaP*ðAÞ¼XB j B#AmðBÞ:ð6ÞIt follows directly from equation(5)P*ðAÞ¼XB j B>A–YmðBÞ;ð7Þfor all A#X.Assume now that evidence is expressed in terms of a given lower probability function P*.Then,the set of probability distribution functions that are consistent with P*,Uncertainty measures on IDM511P ðP *Þ;which is always closed and convex,is defined as followsP ðP *Þ¼p j x [X ;p ðx Þ[½0;1 ;Xx [X p ðx Þ¼1P *ðA Þ<X x [A p ðx Þ;A #X ():ð8Þ2.2Choquet capacities of various ordersA well-defined category of theories of imprecise probabilities is based on Choquet capacities of various orders (Choquet 1953/1954).The most general theory in this category is the theory based on capacities of order 2.Here,the lower and upper probabilities,P *and P *,are monotone measures for whichP *ðA <B Þ>P *ðA ÞþP *ðB Þ2P *ðA >B Þ;P *ðA >B Þ<P *ðA ÞþP *ðB Þ2P *ðA <B Þ;ð9Þfor all A ,B #X .Less general uncertainty theories are then based on capacities of order k .For each k .2,the lower and upper probabilities,P *and P *,satisfy the inequalitiesP *[k j ¼1A j !>X K #N k ;K –Y ð21Þj K jþ1P \j [KA j !;P *\k j ¼1A j!<X K #N k ;K –Y ð21Þj K jþ1P [j [K A j !;ð10Þfor all families of k subsets of X ,where N k ¼{1;2;...;k }:Clearly,if k 0.k ,then the theory based on capacities of order k 0is less general than the one based on capacities of order k .The least general of all these theories is the one in which the inequalities are required to hold for all k >2(the underlying capacity is said to be of order 1).This theory,which was extensively developed by Shafer (1976),is usually referred to as evidence theory or DST.In this theory,lower and upper probabilities are called belief and plausibility measures,noted asBel and Pl,respectively.An important feature of DST is that the Mo¨bius representation of evidence m (usually called a basic probability assignment function in this theory)is a non-negative function (m (A )[[0,1]).Hence,we can obtain Bel and Pl function from m as the following wayBel ðA Þ¼X B j B #A m ðB Þ;Pl ðA Þ¼X B j B >A –Ym ðB Þ:ð11ÞDST is thus closely connected with the theory of random sets (Molchanov 2004).When we work with nested families of focal elements,we obtain a theory of graded possibilities,which is a generalization of classical possibility theory (De Cooman 1997,Klir 2006).2.3Probability intervalsIn this theory,lower and upper probabilities P *and P *are determined for all sets A #X by intervals [l (x ),u (x )]of probabilities on singletons (x [X ).Clearly,l ðx Þ¼P *ð{x }ÞandJ.Abella´n 512u ðx Þ¼P *ð{x }Þand inequalities (1)must be satisfied.Each given set of probability intervals I ¼{½l ðx Þ;u ðx Þ j x [X }is associated with a credal set,P ðI Þ;of probability distribution functions,p ,defined as followsP ðI Þ¼p j x [X ;p ðx Þ[½l ðx Þ;u ðx Þ ;X x [Xp ðx Þ¼1():ð12ÞSets defined in this way are clearly special cases of sets defined by equation (8).Their special feature is that they always form an (n 21)-dimensional polyhedron,where n ¼j X j :In general,the polyhedron may have c vertices (corners),wheren <c <n ðn 21Þ;and each probability distribution function contained in the set can be expressed as a linearcombination of these vertices (Weichselberger and Po¨hlmann 1990,de Campos et al.1994).A given set I of probability intervals may be such that some combinations of values taken from the intervals do not correspond to any probability distribution function.This indicates that the intervals are unnecessarily broad.To avoid this deficiency,the concept of reachability was introduced in the theory (Campos et al.1994).A given set I is called reachable (or feasible)if and only if for each x [X and every value v (x )[[l (x ),u (x )]there exists a probability distribution function p for which p ðx Þ¼v ðx Þ:The reachability of any given set I can be easily checked:the set is reachable if and only if it passes the following testsX x [X l ðx Þþu ðy Þ2l ðy Þ<1;;y [X ;Xx [X u ðx Þþl ðy Þ2u ðy Þ>1;;y [X :ð13ÞIf I is not reachable,it can be converted to the set I 0¼{½l 0ðx Þ;u 0ðx Þ j x [X }of reachable intervals by the formulae l 0ðx Þ¼max l ðx Þ;12X y –x u ðy Þ();u 0ðx Þ¼min u ðx Þ;12Xy –x l ðy Þ();ð14Þfor all x [X .Given a reachable set I of probability intervals,the lower and upper probabilities are determined for each A #X by the formulae P *ðA Þ¼maxX x [A l ðx Þ;12X x ÓA u ðx Þ();P *ðA Þ¼min Xx [A u ðx Þ;12X x ÓA l ðx Þ():ð15ÞThe theory based on reachable probability intervals and DST are not comparable in terms of their generalities.However,they both are subsumed under a theory based on Choquet capacities of order 2as we can see in the following subsection.Uncertainty measures on IDM 5132.4Choquet capacities of order 2Although Choquet capacities of order 2do not capture all credal sets,they subsume all the other special uncertainty theories that are examined in this paper.They are thus quite general.Their significance is that they are computationally easier to handle than arbitrary credal sets.In particular,it is easier to compute P ðP *Þdefined by equation (8)when P *is a Choquet capacity of order 2.Let X ¼{x 1;x 2;...;x n }and let s ¼ðs ðx 1Þ;s ðx 2Þ;...;s ðx n ÞÞdenote a permutation bywhich elements of X are reordered.Then,it is established (de Campos and Bolan˜os 1989)that for any given Choquet capacity of order 2,P ðP *Þis determined by its extreme points,which are probability distributions p s computed as followsp s ðs ðx 1ÞÞ¼P *ð{s ðx 1Þ}Þ;p s ðs ðx 2ÞÞ¼P *ð{s ðx 1Þ;s ðx 2Þ}Þ2P *ð{s ðx 1Þ}Þ;.........p s ðs ðx n ÞÞ¼P *ð{s ðx 1Þ;...;s ðx n Þ}Þ2P *ð{s ðx 1Þ;...;s ðx n 21Þ}Þ:ð16ÞEach permutation defines an extreme point of P ðP *Þ;but different permutations can give rise to the same point.The set of distinct probability distributions p s is often called an interaction representation of P *(Grabisch2000).Figure 1.Main uncertainty theories ordered by their generalities.J.Abella´n 514Uncertainty measures on IDM515 Belief functions and reachable probability intervals represent special types of capacities of order2,as we can see in Figure1.However,belief functions are not generalizations of reachable probability intervals and the inverse is also not verified as we can see in Examples 1and2,respectively:Example1.We consider the set X¼{x1;x2;x3}and the following set of probability intervals on XL¼{½0;0:5 ;½0;0:5 ;½0;0:5 }:This set of probability intervals L has associated a credal set,P L;with vertices{ð0:5;0:5;0Þ;ð0:5;0;0:5Þ;ð0;0:5;0:5Þ}:There does not exist any basic probability assignment for this credal set.To prove this we suppose the contrary condition.Using equation(16)it can be proved that the credal set associated with a basic probability assignment on X has the vertices that we can see in Table1,where m i¼mð{x i}Þ;m ij¼mð{x i;x j}Þ;m123¼mðXÞ;i;j[{1;2;3}:Then,a basic probability assignment m with the same credal set,P L;must verify thatm1þm12þm13þm123¼0:5;m2þm12þm23þm123¼0:5;m3þm13þm23þm123¼0:5;m1¼m2¼m3¼0;m2þm23¼0;m3þm23¼0;m1þm13¼0;m3þm13¼0;m1þm12¼0;m2þm12¼0;where any other option give us a contradiction.Hence,we have that m i¼0;m ij¼0(i,j[{1,2,3})and m123¼0:5;implying that m is not a basic probability assignment.Example2.We consider the following basic probability assignment m on thefinite set X¼{x1;x2;x3;x4}defined bymð{x1;x2}Þ¼0:5;mð{x3;x4}Þ¼0:5:Table1.Set of vertices associated with a basic probability assignment on a set of3elements.s p1p2p3(1,2,3)m1þm12þm13þm123m2þm23m3(1,3,2)m1þm12þm13þm123m2m3þm23(2,1,3)m1þm13m2þm12þm23þm123m3(2,3,1)m1m2þm12þm23þm123m3þm13(3,1,2)m1þm12m2m3þm13þm23þm123 (3,2,1)m1m2þm12m3þm13þm23þm123Computing the upper and lower probability values for every x i ,we have the following set of probability intervals compatible with m :L ¼{½0;0:5 ;½0;0:5 ;½0;0:5 ;½0;0:5 };but this set contains the following probability distribution p 0¼ð0:5;0:5;0;0Þon X ,that not belongs to the credal set associated with m0¼p 0ð{x 3;x 4}Þ,Bel ð{x 3;x 4}Þ¼0:5;1¼p 0ð{x 1;x 2}Þ.Pl ð{x 1;x 2}Þ¼0:5:However,it is easy to obtain a set of reachable probability intervals that represents the same credal set that a belief function,as we can see in the following example.Example 3.We consider the set X ¼{x 1;x 2;x 3}and the following set of reachable probability intervals on XL ¼{½0:3;0:65 ;½0:2;0:55 ;½0:15;0:3 }:This set of probability intervals L has associated a credal set,P L ;with vertices{ð0:65;0:2;0:15Þ;ð0:3;0:55;0:15Þ;ð0:5;0:2;0:3Þð0:3;0:4;0:3Þ}:Using Table 1,it can be obtained that also this credal set is represented by the belief function associated with the basic probability assignment (has the same set of vertices)m ð{x 1}Þ¼0:3;m ð{x 2}Þ¼0:2;m ð{x 3}Þ¼0:15;m ð{x 1;x 2}Þ¼0:2;m ð{x 1;x 2;x 3}Þ¼0:15:3.IDM probability intervalsThe IDM was introduced by Walley (1996)to draw an inference about the probability distribution of a categorical variable.Let us assume that Z is a variable taking values on a finite set X and that we have a sample of size N of independent and identically distributed outcomes of Z .If we want to estimate the probabilities,u x ¼p ðx Þ;with which Z takes its values,a common Bayesian procedure consists in assuming a prior Dirichlet distribution for the parameter vector (u x )x [X ,and then taking the posterior expectation of the parameters given the sample.The Dirichlet distribution depends on the parameters s ,a positive real value,and t ,a vector of positive real numbers t ¼ðt x Þx [X ;verifying P x [X t x ¼1:The density takes the formf ððu x Þx [X Þ¼G ðs ÞQ x [X G ðs ·t x ÞY x [Xu s ·t x 21x ;where G is the gamma function.If r (x )is the number of occurrences of value x in the sample,the expected posterior value of parameter u x is (r (x )þs ·t x )/(N þs ),which is also the Bayesian estimate of u x (under quadratic loss).J.Abella´n 516The IDM(Walley1996)only depends on parameter s and assumes all the possible values of t.This defines a non-closed convex set of prior distributions.It represents a much weaker assumption than a precise prior model,but it is possible to make useful inferences.In our particular case,where the IDM is applied to a single variable,we obtain a credal set for this variable Z that can be represented by a system of probability intervals.For each parameter, u x,we obtain a probability interval given by the lower and upper posterior expected values of the parameter given the sample.These intervals can be easily computed and are given by [r(x)/(Nþs),(r(x)þs)/(Nþs)].The associated credal set on X is given by all the probability distributions p0on X,such that p0(x)[[r(x)/(Nþs),(r(x)þs)/(Nþs)],;x.The intervals are coherent in the sense that if they are computed by taking infimum and supremum in the credal set,then the same set of intervals is again obtained.Parameter s determines how quickly the lower and upper probabilities converge as more data become available;larger values of s produce more cautious inferences.Walley(1996) does not give a definitive recommendation,but he advocates values between s¼1and s¼2. We can define a generalization of a set of IDM probability intervals,considering that the frequencies r(x i)are non-negative real numbers.For the sake of simplicity,we use the same name for this type of probability interval.Formally:Definition1.Let X¼{x1;...;x n}be afinite set.Then a set of IDM probability intervals on X can be defined as the setL¼½l i;u i j l i¼rðx iÞNþs;u i¼rðx iÞþsNþs;i¼1;2;...;n;X ni¼1rðx iÞ¼N();where r(x i)are non-negative numbers and not all are equal to zero,and s is non-negative parameter.3.1PropertiesUsing the notation in definition1,we can express the following properties:1.Sets of IDM probability intervals generalize probability distributions.For a probabilitydistribution p on afinite set X¼{x1;...;x n};it is only necessary to consider s¼0and rðx iÞ¼pð{x i}Þ;for all i¼1;...;n:2.The credal set associated with a set L of IDM probability intervals,P L;has the followingset of vertices{v1,...,v n}v1¼rðx1ÞþsNþs;rðx2ÞNþs;...;rðx nÞNþsv2¼rðx1ÞNþs;rðx2ÞþsNþs;...;rðx nÞNþs.........v n¼rðx1ÞNþs;rðx2ÞNþs;...;rðx nÞþsNþsð17ÞUncertainty measures on IDM5173.Denoting as P s L the credal set associated with a set L of IDM probability intervals for a valueof the parameter s and a fixed array of values r ¼ðr ðx 1Þ;...;r ðx n ÞÞ;it can be verified thats 1<s 2,P s 1L #P s 2L4.Every set of IDM probability intervals represents a set of reachable probability intervals.InSection 2.4,we see that belief functions are not generalizations of probability intervals and the inverse is also not verified.However,the credal set associated with a set of IDM probability intervals L can also be expressed by a belief function.Proposition 1.Let L be a set of IDM probability intervals as in Definition 1.The credal set associated with L is the credal set associated with the belief function associated with the basic probability assignment m Lm L ð{x i }Þ¼r ðx i ÞN þs ;i ¼1;2;...;n m L ðX Þ¼sN þsm L ðA Þ¼0;;A ,X ;1,j A j ,n :ð18ÞProof .Using that the lower probability associated with L verifies thatP *ð{x i ;...;x j }Þ¼r ðx i Þþ···þr ðx j ÞN þs;via the Mo¨bius transform,we can obtain the following values m L ð{x i }Þ¼r ðx i ÞN þs ;m L ð{x i ;x j }Þ¼r ðx i Þþr ðx j ÞN þs 2r ðx i ÞN þs 2r ðx j ÞN þs ¼0;m L ð{x i ;x j ;x k }Þ¼r ðx i Þþr ðx j Þþr ðx k ÞN þs 2r ðx i Þþr ðx j ÞN þs 2r ðx j Þþr ðx k ÞN þs 2r ðx j Þþr ðx k ÞN þs þr ðx i ÞN þs þr ðx j ÞN þs þr ðx k ÞN þs ¼0;......;ð19Þfor all i ,j ,k [{1,2,...,n }.For a general set A such that 1,j A j ¼w ,n ;we have m L ð{A }Þ¼X B #A ð21Þj A 2B j P *ðB Þ¼X B #A ð21Þj A 2B j P x i [B r ðx i ÞN þs ¼X x i [Aw 2100@1A 2w 2210@1A þw 2330@1A 2···þð21Þw 21w 21w 210@1A 2435:r ðx i ÞN þs :ð20ÞJ.Abella´n 518Taking into account that0¼ð121Þw21¼w21!2w221!þw233!2···þð21Þw21w21w21!;ð21Þthenm Lð{A}Þ¼0;ð22ÞNowm Lð{X}Þ¼12Xx i[Xrðx iÞNþs¼sNþs:ð23ÞTherefore,m L obtained is a basic probability assignment on X.Now,let P L be the credal set associated with L and let P mLbe the credal set associated with m L.Then,P L¼P mL:i)Let p[P L be a probability distribution.ThenBel mL ðAÞ¼Xx i[Arðx iÞNþs<pðAÞ<Px i[Arðx iÞþsNþs¼Pl mLðAÞ;for all A#X.Hence,p[P mL;ii)Let p[P mLbe a probability distribution.Thenrðx iÞNþs ¼Bel mLð{x i}Þ<pð{x i}Þ<Pl mLð{x i}Þ¼rðx iÞþsNþs;for all x i[X.Hence,p[P L:A Sets of IDM probability intervals are not the only credal sets that can be expressed jointly by reachable probability intervals and belief functions.As we can observe in example3,it is possible for a credal set to be represented by a set of reachable probability intervals and by a belief function,although this credal set cannot be represented by a set of IDM probability intervals.We only need to consider in example3the value s/(Nþs):it must be0.35,using l1 and u1and0.15using l3and u3.However,the description of the credal sets belonging to reachable probability intervals and belief functions is still an open problem.In Figure1,we can see where the sets of IDM probability intervals are placed in relation to other theories of imprecise probabilities using a generality order.4.An overview of uncertainty measuresIt has well been established that uncertainty in classical possibility theory is quantified by the Hartley measure(Hartley1928).For each nonempty andfinite set A#X of possible alternatives,the Hartley measure,H(A),is defined by the formulaHðAÞ¼log2j A j;ð24ÞUncertainty measures on IDM519。

物理不可克隆函数熵源

物理不可克隆函数熵源

物理不可克隆函数熵源
物理不可克隆函数熵源(P-Uncloneable Entropy Sources,P-UES)是一种由物理
设备提供的不可克隆的随机性。

它们将低成本的物理噪声转换成大量的原始、安全的随机性。

在这个过程中,传感器和处理器是必不可少的因素。

传感器从物理环境中获取信号,而处理器利用收集到的信号来生成真正随机性并将它作为加密密钥使用。

传感包括光学、声学、磁学、流动学、生化学等方面。

常见的光学传感器包括图像传感器,如CMOS或CCD传感器;声音传感器包括话筒或扬声器;磁场传感包括
金氧半元件或敏感度设备。

而处理部分则包含了通过软件整合各项信号并把它们合并生成一个真正安全的加密密钥来保证通信中数据的安全性。

此外,该物理不可克隆函数熵源也得到了应用,如生物语验、金融服务、行动装置中使用交易核心来保证交易安全. 同时在交易核心凭证已无法重新使用一遍;圈子
间交流中也使用到此函数决不通途径再利用凭证内容. 乘客使用时乘客核心上乘客
凭证已无法再使用一遍; 电子市集场所中能够通过此函数来防止市集者之间盗版行办. 由上述行办能看到, 近年来, 物理不可克隆性函数已然得到大量应用.
总之, 物理不可克隆函数熵源是一种低成本而强大的方式来生产原始安全的随机性。

通过对低成本物理信号进行测量并将其汇总为一个真正安全的加密密钥,物理不可克隆函数熵源为通信过程中提供了坚固的安全保障。

gentle代数的模范畴和导出范畴—曲面的组合与稳定条件

gentle代数的模范畴和导出范畴—曲面的组合与稳定条件

gentle代数的模范畴和导出范畴—曲面的组合与稳定条件在数学中,gentle代数是一种特殊的代数结构,可以用来描述有限维向量空间和矢量的组合法则。

模范畴和导出范畴是与gentle代数相关的两个重要的范畴。

模范畴是将gentle代数的结构推广到更一般的范畴情形的概念。

具体来说,给定一个gentle代数A,模范畴Mod(A)定义为所有A-代数的集合。

其中,A-代数是一个结合性的代数结构,包含一个对象和一个从对象到自身的映射,满足一定的性质。

模范畴中的箭头是A-代数之间的映射。

模范畴是一个重要的研究对象,因为它能够描述gentle代数的一些重要的性质和结构。

导出范畴是一种用来描述gentle代数之间关系的范畴。

具体来说,给定两个gentle代数A和B,导出范畴Der(A,B)定义了一种从A到B的箭头的集合。

箭头通常被称为导出映射,它们满足一定的性质。

导出范畴是研究gentle代数之间的映射关系和变换的有用工具。

曲面的组合是gentle代数中的一个重要概念。

它指的是将两个gentle代数A和B通过组合操作得到一个新的gentle代数C的过程。

具体来说,如果A和B是两个gentle代数,那么它们的组合A∘B可以定义为A和B的笛卡尔积的张量积,然后对张量积进行一系列的运算得到结果。

曲面的组合在gentle代数中是非常重要的,因为它可以用来描述代数结构之间的关系和变换。

稳定条件是gentle代数中的一个重要的性质。

一个gentle代数被称为是稳定的,如果它满足一定的条件。

具体来说,稳定条件指的是gentle代数中定义的一些操作和算法在特定的条件下是收敛的。

稳定条件是gentle代数中的一个基本概念,它能够保证gentle代数的一些重要的性质和结构的存在。

总之,gentle代数的模范畴和导出范畴、曲面的组合以及稳定条件是gentle代数理论中的重要概念和工具,它们被广泛应用于代数结构的研究和应用。

基于Arnold变换和混沌映射的图像加密方法

基于Arnold变换和混沌映射的图像加密方法

作者: 邢顺来[1] 李志斌[2] 周华成[3]
作者机构: [1]华东师范大学计算机系 [2]济南大学数学科学学院 [3]东华大学应用数学系出版物刊名: 山东广播电视大学学报
页码: 42-45页
年卷期: 2012年 第1期
主题词: 图像加密 混沌映射 Arnold变换
摘要:在研究现有的图像加密算法及混沌序列的相关知识的基础上,提出一种基于Arnold变换和混沌映射的图像加密方法。

通过二维Arnold变换改变图像像素的位置,继而用三维Arnold 变换改变像素的值并引入混沌映射再次改变像素的值对一幅真彩色图像加密;再结合图像特征,把图像分割成几个区域,对每一个区域用同样的方法实现,从而实现图像加密的可靠性。

实验证明了图像加密解密的有效性,表明该方法加密、解密效果更好。

基于混沌映射的多因子认证密钥协商协议

基于混沌映射的多因子认证密钥协商协议

基于混沌映射的多因子认证密钥协商协议王松伟;陈建华【期刊名称】《计算机应用》【年(卷),期】2018(038)010【摘要】在开放的网络环境中,身份认证是确保信息安全的一种重要手段.针对Li 等(LI X,WU F,KHAN M K,et al.A secure chaotic map-based remote authentication scheme for telecare medicine information systems.Future Generation Computer Systems,2017,84:149-159.)提出的身份认证协议,指出其容易遭受用户冒充攻击、拒绝服务攻击等缺陷,并提出一个新的多因子认证协议来修复以上安全漏洞.该协议使用了扩展混沌映射,采用动态身份保护用户匿名性,并利用三次握手技术实现异步认证.安全性分析结果表明,所提协议可以抵抗冒充攻击、拒绝服务攻击,能够保护用户匿名性和身份唯一性.【总页数】6页(P2940-2944,2954)【作者】王松伟;陈建华【作者单位】武汉大学数学与统计学院,武汉430072;武汉大学数学与统计学院,武汉430072【正文语种】中文【中图分类】TN918【相关文献】1.基于混沌映射的用户匿名三方口令认证密钥协商协议 [J], 王彩芬;陈丽;刘超;乔慧;王欢2.WSNs中一种基于Chebyshev混沌映射的认证密钥协商协议 [J], 郭琰;石飞;汪烈军;刘双3.基于扩展混沌映射的三方认证密钥协商协议 [J], 闫丽丽;昌燕;张仕斌4.基于智能卡的扩展混沌映射异步认证密钥协商协议 [J], 王松伟;陈建华5.基于扩展混沌映射的动态身份认证密钥协商协议 [J], 曹阳因版权原因,仅展示原文概要,查看原文内容请购买。

  1. 1、下载文档前请自行甄别文档内容的完整性,平台不提供额外的编辑、内容补充、找答案等附加服务。
  2. 2、"仅部分预览"的文档,不可在线预览部分如存在完整性等问题,可反馈申请退款(可完整预览的文档不适用该条件!)。
  3. 3、如文档侵犯您的权益,请联系客服反馈,我们会尽快为您处理(人工客服工作时间:9:00-18:30)。

a r X i v :0803.3918v 2 [c o n d -m a t .d i s -n n ] 29 A p r 2008Universality class of replica symmetry breaking,scaling behavior,and the low-temperature fixed-point order function of the Sherrington-Kirkpatrick modelR.Oppermann,M.J.SchmidtInstitut f.Theoretische Physik,Universit¨a t W¨u rzburg,Am Hubland,97074W¨u rzburg,FRG(Dated:April 30,2008)A scaling theory of replica symmetry breaking (RSB)in the SK-model is presented in the frame-work of critical phenomena for the scaling regime of small inverse RSB-orders 1/κ,small tempera-tures T ,and small magnetic fields H .We employ the pseudo-dynamical picture (Phys.Rev.Lett.98,127201(2007))with two critical points CP 1and CP 2where separated temperature-and mag-netic field-scaling is obtained near the order function’s pseudo-dynamical limits lim a →∞q (a )=1and lim a →0q (a )=0at (T =0,H =0).An unconventional scaling hypothesis for the free energy is given,modeling this separated scaling in accordance with detailed numerical self-consistent solutions for up to 200orders of RSB.Divergent correlation-lengths ξCP 1(T )∼T −νT and ξCP 2(H )∼H −νH describe the RSB-criticality as a long-range correlation effect occurring on the pseudo-lattice of RSB-orders.Rational-valued exponents νT =3/5and νH =2/3are concluded with high precision from high-order RSB scaling (in analogy with finite size scaling)and using a new fixed point extrap-olation method.Power laws,scaling relations,and scaling functions are analyzed.Near CP 1,the non-equilibrium susceptibility is found to decay like χ1=κ−5/3f 1(T /κ−5/3),the T =0-entropy likeS ∼χ21,while (T -normalized)Parisi box sizes diverge like a i =κ5/3f a i (T /κ−5/3),with f 1(ζ)∼ζand f a i (ζ)∼1/ζfor ζ→∞,f (0)finite.Near CP 2,where the magnetic field H controls the critical behavior (while temperature is irrelevant),a power law H 2/3is retrieved for plateau-height of the order function q (a )according to q pl (H )=κ−1f pl (H 2/3/κ−1)with f pl (ζ)|ζ→∞∼ζand f pl (0)finite.The order function q (a )links CP 1with CP 2and is obtained as a fixed point function q ∗(a ∗)of RSB-flow,in agreement with integrated fixed-point energy and susceptibility distributions.Similarities with directed polymers in 1+1dimensions,with d =1solution and Flory-Imry-Ma type solutions of the KPZ-equation are discussed.PACS numbers:75.10.Nr,75.40.-s,89.75.DaI.INTRODUCTIONThe far-reaching usefulness of spin glass theories 1,2,3and of its key structural elements such as frustration,disorder,hierarchical order,ultrametricity,complexity,freezing transition,is witnessed well by applications en-tering even life-sciences and trans-disciplinary research fields.Physical models,where these key structures ac-quired a specific mathematical meaning,find very broad applications beyond their origin in frustrated magnetism.Let us mention,apart from fields like neural networks,computer science,and econophysics,the fascinating so-ciological applications to opinion-and group dynamics 4,5,biological applications to RNA-folding 6,7,8including the quantum chromodynamical analogy and random matrix theory 9.It seems natural to search for universal features of unifying models both in the general sense and in the precise meaning of the renormalization group 7.The 3SAT optimization problem and its close rela-tion with the T =0Sherrington-Kirkpatrick model 12or RNA-folding in biophysics 6,where glass transitions exist within the secondary RNA-structure 7,10,11,provide ex-amples where even the zero temperature limit is either exact or close to the realistic situation.In physics,spin glass phases are usually confined to a low temperature regime and some applications are rather remote from it.Yet,knowing the ground state structure remains impor-tant.For one of the most fertile standard models,theSherrington-Kirkpatrick model 12(SK-model),the hier-archical ground state structure is meanwhile confirmed 13as predicted by Parisi a long time ago 14,15.Explicit analytic solutions on the other hand or meaningful ap-proximations are still required.They may lead to im-proved understanding and could be potentially fruitful for progress in more complicated (non-mean-field finite-range,quantum-)models.The attempt to link the SK-model behavior deep in-side its ordered (spin glass)phase with the theory of crit-ical phenomena may appear unmotivated at first sight,since the infinite-ranged spin interaction suggests ’only’mean-field behavior.However the SK-model is not simple below its mean-field transition.Its replica theory 16al-lows to imagine how the Ising Hamiltonian with infinite-ranged random interaction can become potentially criti-cal,when it is dressed-up with the hierarchical order pa-rameter structure in the replica symmetry broken (RSB)phase 3.We shall argue in this paper that,as the number of tree levels of this hierarchical structure grow to infin-ity,a particular correlation length between the tree levels can be defined which diverges as T →0.It allows to de-scribe critical behavior due to the accumulated effect of ever finer structures at the highest tree level.This prop-erty specifies a kind of universality class,which helps to compare with similar behavior in different physical sys-tems and in other scientific areas like biology,sociology,(mathematical)psychology as well,where evidently frus-trated random (and in cases range-free)interactions are2important.Nonanalytic power laws(with rational exponents)for the SK-model had been discussed in many different re-spects,as for example thefinite-size cutoff(orfinite spin number)dependence17,18,19.One may also mention the exponent of the Almeida Thouless line16.However a link to specific critical points was not made.In the present paper,we shall report progress in under-standing replica symmetry breaking in the Sherrington-Kirkpatrick model12as a critical phenomenon;this refers to scaling behavior on one hand and to(numerically de-termined)fixed point functions under RSB-orderflow κ→∞on the other.Nonanalytic scaling behavior is described as a function of the inverse RSB-order decreas-ing to zero either together with temperature T→0,or together with the external magneticfield.Temperature-andfield-scaling are well separated and reside in opposite limits of a pseudo-dynamic variable1/a(see Ref.20),as sketched by Fig.153.Critical phenomena are in general categorized by uni-versality classes and described by criteria like global sym-metries.Certain details(on shorter range)become ir-relevant and suppressed in the regime of divergent cor-relation lengths.In the early years of the develop-ment of phase transition theory and critical phenomena, Kadanoff’s initializing ideas of universality and rescal-ing,Stanley’s scaling theory,and Wilson’s renormaliza-tion group led to the modern understanding of critical behavior21.In recent years the functional renormaliza-tion group was advocated to understand better disorder-related criticality22.Freezing transitions into spin glass phases were an-alyzed in renormalization framework too,the ordered phase itself remained however mysterious,in particular for the non-mean-field models.In a famous work on scaling in spin glasses D.Fisher and Sompolinsky23ex-plained the complications of meanfield models(or mean field regimes offinite-range spin glasses above d=6and d=8)and the multiple violations of scaling relations.In particular they mentioned the violation of temperature-versus magneticfield scaling within the ordered phase. In a different manner,we re-encounter this problem and explain a certain decoupling offield-from temperature-scaling by the presence of two different critical points of RSB in the low temperature limit.Crucial questions like the relationship between Parisi’s RSB and the Fisher-Huse droplet theory24of the ordered phase of real spin glasses(or their reconciliation)became -since a long time but perhaps currently with more good reason-afield of intensive research25,26.Since droplet theory is interpreted to govern the ordered phase by a T=0fixed point,it appears very desirable to understand RSB as a T=0fixed point theory too.Attempts have been independently undertaken by several authors and also in differentfields of application,as the examples in Refs.20,22,27,28,29show.The latter point is elaborated in the present article. Despite the meanfield character of the SK-model,RSB introduces apparently nonanalytic critical behavior of one-dimensional type(the unbroken replica-symmetric solution does not show any of these phenomena)together with special diverging correlation lengths.The challenge to handle RSB-effects correctly and to make the SK-solution a fruitful basis for real physical applications led us to a scaling theory intimately based on extreme high order numerical results.In previous publications20,27we reported the existence of two critical points and of discrete spectra which sur-vived in the limit of infinite replica symmetry breaking (∞-RSB)for the SK-model at T=0,perhaps surpris-ing since the∞-RSB limit is in generally known only as the’continuum limit’.Indeed,a continuum scaling theory,dealing with the T→0−limit atκ=∞)was published by Pankov28recently.Its role and limitation to the temperature-controlled critical point CP1has been addressed in our previous publication30together with a comparison of our work with the much older so-called PaT-scaling31.In the present article we do neither use Pankov-nor PaT-scaling,but construct a different scal-ing approach,which includes RSB-order-scaling,and is exclusively guided by the theory of critical phenomena. In accordance with previous(naive)functional renormal-ization group arguments27we analyze the approach to full RSB formation(κ→∞)not only at T=0but also in the(H,T)-plane for small values of temperature T and magneticfield H and,of course,as a function of RSB-order.(neither real space nor real-time space are involved as a consequence of the SK-model’s nature). We suppose here that RSB orders,counted by inte-gers1,2,...,κ,can be viewed as equidistant sites form-ing a pseudo-lattice.In analogy with a real-space lat-tice,which needs to be infinitely large in order to allow diverging correlation lengths and hence support critical phenomena,the pseudo-length cutoffκmust be sent to infinity.The known fact that increasingly high orders of RSB are needed(for good approximations)as the tem-perature decreases towards zero implies the role of T as an effective cutoffof nonanalytic behavior in the RSB-limit(T playing the role of a symmetry breaking relevant perturbation in standard critical phenomena).Thus it also inherits the idea of scaling RSB-orderκwith tem-perature T.Conversely,a maximum RSB orderκserves as a cutoffof criticality.A speciality of RSB is that it appears in the shape of a pseudo-dynamical critical phenomenon20,27,which recalls the celebrated dynamical representation of Sompolinsky32.A technically impor-tant difference however being the absence of a stochastic field,which we reserve for more complicated couplings to faster degrees of freedom33.A scaling theory,near T=0in particular,is im-portant for several different reasons.First,it expresses the numerically determined features of the SK-model in a universal form,which helps to identify model-independent features and places the SK-model and its RSB in a wider context.Let us mention that directed polymers(or for example the queuing transition and3 the totally asymmetric exclusion process34,35,or certainpartial differential equations,involve rational exponentsas multiples of1/3too).The scaling theory also putsconstraints on the shape of an effectivefield theory.Ithas the virtue of isolating critical features which mustbe represented correctly by an effective theory thatsimplifies the SK-model.54The simpler theory shouldallow to control generalizations tofinite range or othercomplications.The scaling theory offers also a special look on eventual scenarios of an RSB breakdown,as it may occur due tofinite range interactions.The collapse offinite T c below a lower critical dimension, will eventually combine RSB-criticality with the freezing transition as T c=0.The paper is organized as follows:SectionII and III describe the basic elements of the present scaling theory.The spaces spanned by the scaling variables at both critical points are described in section II.In section III a correlation length is introduced on the pseudo-lattice of RSB-orders(to our best knowledge for thefirst time)and,anticipating the self-consistent numerical results of the following Sections IV-VIII(for details see Ref.30),the role offinite temperatures(or finite magneticfields)as soft cut-offs of the divergence of this correlation length is explained.Section IV demonstrates how the order parameter function q(a)can be regarded and obtained as afixed point function q∗(a∗)under RSB-flowκ→∞.Section V includes and combinesfinite temperature scaling near critical point CP1with RSB-order scaling. Scaling functions are obtained,whichfit the detailed data of200RSB-orders,and explain the non-commuting sin-gular limitκ→∞,T→0.In a similar way,VI includes magneticfield scaling near CP2.In Section VII we present unconventional scaling-contributions to the free energy,to the entropy,and in-ternal energy,which are compatible with the numerical self-consistent solutions.In Section VIII the ground state energy distribution is given as a function of pseudo-time and also shown as a function of the(normalized)Parisi levels l/κsuch that theflow towards a energy-per-levelfix point function re-sults as the RSB-order tends to infinity.In Section IX wefinally consider pseudo-dynamic scal-ing of the order function q(a)in the vicinity of both crit-ical points before concluding with details of q(a)as re-vealed by its derivatives in Section X.II.THE SCALING SCENARIOWe introduce the(RSB-)scaling idea by viewing the formation of full RSB as a critical phenomenon with two critical points in the pseudo-dynamic limits a=0and a=∞at T=0,H=0.We do not a priori impose a relationship between the two critical points,but consider the pseudo-dynamical crossover between them by means of the order function q(a)on0≤a≤∞.Fig.1illustrates the relative position of the two critical points and the sets of scaling variables near these points.In particular one may notice that the dynamical vari-able1/a and the RSB-orderκdefine a(1+1)-dimensional analogy of problems with one time-and one real space dimension.Since the free energy or internal energy are FIG.1:Scaling regimes near the two critical points of the SK-model at zero temperature,zerofield,and infinite replica symmetry breaking.The critical points are separated by the full range of continuously distributed pseudo-times1/a. integrals over all pseudo-times,as for example given be-low in Eqs.(24),(26),we do not start from a single scaling hypothesis for the free energy F.Instead we construct the scaling hypothesis for each of the two different scal-ing contributions,originating in these separated critical points and.As Fig.1shows,a different set of scaling variables should be used in order to match the numerical results.It is remarkable that temperature-and magneticfield-scaling become decoupled,because they belong to differ-ent scaling regimes.Scaling with the respect to the order κof RSB measures the approach of the equilibrium so-lution atκ=∞(full RSB)and therefore can be viewed as a kind of non-equilibrium dynamics(in the sense that eachfinite order is unstable towards higher RSB-orders). Thus an element of dynamic scaling is ing the pseudo-time1/a as an additional scaling variable, we analyze the order function q(a;T,H)and its pseudo-dynamic scaling behavior.A dynamic crossover between the two critical points CP1and CP2is then described by means of q(a).Moreover,the order function is evaluated as afixed point function of the RSB-flow lettingκ→∞. The present scaling theory is thenfitted to high preci-sion numerical data,which were obtained recently for the Sherrington-Kirkpatrick model given by the HamiltonianH= i<j J ij s i s j−H i s iwith quenched,infinite-ranged,and Gaussian-distributed random couplings J ij(with variance J2/N)between clas-sical spins s i=±1.The method was described in Ref.30 and will not be described again in this article.It al-lowed not only to go beyond earlier high-order studies20, but contained also new analytical elements.As a conse-quence we are able to predict the values of critical expo-nents,evaluate amplitudes,calculate analytical models of various scaling functions including cases with very sin-gular crossover.The numerical material includes the self-consistent solutions in all orders of RSB up toi)the current maximum ofκ=200RSB at T=0and H=0,ii)50orders for a dense grid offinite temperatures in the range0≤T≤0.3for H=0,andiii)20orders of RSB for a dense grid offinite magnetic fields0≤H≤0.5at zero temperature.We note that all energies are given in units of J.III.DIVERGENT CORRELATION LENGTH ON THE ONE-DIMENSIONAL PSEUDO-LATTICEOF RSB-ORDERSThe high order self-consistent solutions led us to con-sider a pseudo-lattice of RSB-orders with unit lattice con-stant.The maximal orderκ,for which self-consistent re-sults have been obtained numerically,can be viewed as a (sharp)pseudo-length cutoff.In analogy with the well-knownfinite size scaling of critical phenomena,one may consider scaling by varying thisfinite maximum RSB-orderκ.Naturally this defines a one-dimensional prob-lem without translational invariance though.Moreover it is known thatfinite temperatures orfinitefields serve each as a soft-cutofffor the maximal order of RSB,which is needed to obtain good approximations:the higher the temperature the less RSB orders are needed to obtain a certain quality of results55.In other words,higher orders become uncorrelated be-cause they only have weak and/or negligible effects.In this sense the correlation length of different RSB or-ders becomes cutoffbyfinite T or H.Anticipating our results below,this definition of the correlation lengths ξκ(a,T,H)shows power law divergences with rational-valued exponents given byξCP1≡ξκ(a=∞,H=0,T)∼T−νT,νT=3/5(1)ξCP2≡ξκ(a=0,H,T=0)∼H−νH,νH=2/3(2) In the chapters below,we shallfind scaling functions of the form f(κ/ξκ).Apparently the correlation length exponentνH cannot be simply related toνT,since in the vicinity of one of the two critical points onefinds either T and no H-dependence(CP1)or vice versa(CP2)(unlike conventional scaling where T∼H1/βδ).The power laws (1)and(2)should also hold for a≫ξ0and a≪ξ0 respectively,whereξ0was introduced in Ref.27as afinite characteristic length(≈1.13)which sets the scale for crossover from almost linear regime,q∼a,through a maximal-curvature crossover(a≈ξ0)to1−q≈1/a2 behavior of the order function q(a,T=0,H=0).Let us note some similarities with critical behavior of other systems.Very remarkable appears the fact that the exponentνT=3FIG.2:Main Figure and Insert present an(overlapping)wide-scale overview of all(T=0,H=0)selfconsistent solutions{a(κ)l ,q(κ)l+1}and{a(κ)l,q(κ)l},where l=1...κfor each RSB-orderκ=1,2,...,200.One color cycle fromκ=1(red,remotepoints)to the highly dense regimeκ=200(red,almost continuous curve)demonstrates how the true order function q(a) is approached by the almost continuous q l+1(a l)from below and by q l(a l)from above.The Insert shows the corresponding {1/a,q}-data and the function q(1/a)they converge too.The shape nearτ≡12 whichfinally turns into q(a)in the RSB-limit.Our pre-viously published analytical model function20satisfies al-most perfectly this constraint;as mentioned in Ref.20it turned out that a small’mass’function w(a)in q model(a)=aa2+w(a)1F1 α,γ,−ξ2FIG.3:Self-consistent{a,q}-data in the small-a regime are shown for allκ=1,2,..200RSB orders,together with an analytical model function q(a)(dashed grey line).So-lutions{a l,κ,q l+1,κ}(above the line)and{a l,κ,q l,κ}(be-low)approachfixed points under constrained RSB-flow obey-ingκ=κ(l),l level number.Fixed point examples for κ=l(m+1)/m−k/m−1,23≥m≥6and k integer are shown(red dots)including thefixed point(0,0)(RSB-flow alongκ=l+k).The analytical model function for q(a) matches well allfixed points.A.Fixed points calculated from the RSB-flowtowardsκ=∞The full set of self-consistent solutions for order pa-rameters q l and(T-normalized)Parisi box sizes a l≡m l(T)/T|T=0can be described by matrix elementsp l,κ≡{a l,κ≡a(κ)l ,q l,κ≡q(κ)l}labeled by RSB-orderκand level number l.Since the number of q l-parameters exceeds by one the number of a l-parameters(in each order of RSB),a second complementary set of matrix elements˜p l,κ≡{a l,κ,q l+1,κ}should also be taken into account.These points p l,κand˜p l,κare displayed in the Figures2−6and observed to approach the exact q(a)≡q∗(a∗)along characteristic lines given below by Eq.(4)asκ→∞(p from above and˜p form below q(a) since q l+1,κ<q l,κ).The set of all RSB-solutions up to a maximum order κ,as plotted in Fig.2with a cutoffatκ=200,is then de-scribed by two triangular matrices with entries{a l,κ,q l,κ} (or with{a l,κ,q l+1,κ});the level-numbers l run from1to κfor each RSB-orderκ.Along infinitely many lines in(l,κ)-space-the leading ones are very clearly visible in Figures3,4(and shown as calculated in Figs.5,6)-we observe very smooth behavior of slowly changing parameters(a l,κ,q l,κ)which allow low order Pad´e-approximants to match these data and to join infixed points p∗of the order function curve forκ=∞.A special case is the origin where thefixed point is obtained with the extreme accuracy of O(10−13)). Typical examples of such characteristic lines in(l,κ)-FIG.4:Two examples of alignment of solutions{a,q} under constrained RSB-flow withκ(l)(or l(κ))towards twofixed points(big dark-red dots):(top)p∗m=7,n=6≡(a∗,q∗)m=7,n=6=(0.3820,0.2976)and(bottom)p∗m=3,n=2≡(a∗,q∗)3,2=(0.9133,0.6559).Pad´e-line intersections,shown for p∗7,6in Fig.5and for a large-afixed point in Fig.6)deter-mine thefixed points.Closer look shows that RSBfix-points lie dense and lead to afixed point function,well approximated by the model function q(a)(dashed grey).space can be given by the linear relation among the labels{l+k,κ=m6l−k/6−1forintegers k=−4,−3,...,4,with initial values l0=6+k,κ0=6+k and∆l=6.All lines join in the RSBfixed pointp∗7,6=(0.3820,0.2976)(red dot)asκ→∞.(Pade curves aredisplayed here without termination at thefix point.Colors dis-FIG.6:Fixed point example in the a≫1-regime is displayedwith nonlinear RSB-flow alongκ=6l+k,k=−2,−1,..,2,modeled by Pade-curves ending in a large-afixed point.Itsposition(big blue dot)at{10.6736,0.9964}and that of a setof50otherfixed points(red dots),along the model functionq(a)(green line),is also shown.order of RSB.B.Discrete spectra in theκ=∞RSB-limit at zerotemperatureWhile thefixed point functions can be derived for allpseudo-time values1/a,the points a=0and a=∞re-main special limits.In a recent article20we have shownthat infinitely large subclasses of certain self-consistentparameters ratios remain discrete at T=0or H=0even in the continuum limit.These discrete levels residein the limits a=0and a=∞whenκ=∞.Finitetemperatures lift the discrete spectrum at a=∞intothe continuum,while a magneticfield has a similar ef-fect on the discrete levels at a=0.The ratios assumethe value1then.The discrete spectra therefore empha-size the critical nature of the points a=0and a=∞.We present in the following subsections new results forthese T=0levels of parameter ratios and,in Section V,describe their singularfinite T crossover.1.Level distribution at CP2At the critical point CP2the sub-class of small self-consistent parameters q k and a k,which vanish in the∞RSB limit(and condense into CP2),obeyq¯l+22l+1anda¯l+1l+1,(5)with¯l≡κ−l and l=1,2,...;thus the ratios of theseparameters are discrete and almost equidistant20.Re-curring these relations to the smallest parameters of eachRSB-orderκ,hence to qκ+1and aκrespectively,we ob-tainqκ+1−l=(2l+1)qκ+1,aκ−l=(l+1)aκ.(6)The RSBflow of numerical data up to200RSB allow toconclude that these minimal parameters vanish likeq min≡qκ+1=1.03059κ2+O(1/κ3),(7)a min≡aκ=2.77275κ2+O(1/κ3).(8)The discretized slope of the order function in the pointa=0,assumes the200RSB valueq¯l−q¯l−1aκ=0.74345(9)or,by Pad´e approximation of the RSBflow and extrap-olation to∞−RSB,one obtainsq′(0)=2limκ→∞qκ+1w(0) =0.743368.2.Level distribution at CP1In the large a limit the characteristic feature are dis-crete spectra of1−q l-ratios,which are shown in Figure87.In addition Figure8shows that the1γ1a2+O(1/a4),(11)putting a constraint onαξ2γ.Further constraints canbe found from very precise numerical characteristics;itis planned to use this analysis to narrow down the choiceof an analytical order function model.The discrete spectrum yields a coefficient which differsnotably from this value,as one can see from Figure7(right)for the leading divergent a l parameters.FIG.7:Thefinite and discrete ratios of1−q l-factors,whichvanish forκ→∞at T=0,as obtained by Pad´e extrapolationto the RSB limit,plotted versus the level number l.C.Approach of equilibrium at T=0:leading andsub-leading scaling contributionsThe nonequilibrium susceptibilityχ1is a characteristicquantity measuring the distance from the equilibrium so-lution atκ=∞.The entropy had been seen30to vanishlike the square ofχ1.The numerical solutions30forχ1,evaluated for all200leading RSB-orders,are wellfittedby the T=0-formχ1(κ,T=0)∼=0.86(κ0+κ)4+ (12)=0.86κ−5/3−1.83κ−8/3+3.12κ−11/3+1.85κ−4+...withκ0∼=1.278.As discussed in Ref.30the numericaluncertainty of O(10−6)in the exponent is so small thatT(13)where we considerfirst scaling in the(κ,T)-plane forfixed label l.We must analyze the singular behaviornear the critical point CP1,where diverging a l(κ,T=0)→∞forκ→∞lead to discretely spaced ratiosa l(∞,0)/a l−1(∞,0)in the∞−RSB limit.We identifiedthe large order power law divergencea l(κ,T=0)∼κ5/3,κ→∞(14)FIG.9:Main Figure shows the RSB-flow of{a l(κ),q l(κ)}(above)and{a l(κ),q l+1(κ)}(below),for each of the six highest parameter levels(l=1...6)atfixed temperature T=0.03.Pad´e approximations(shown for largest level(orange),2nd-largest (green)etcetera)model theflowκ→∞by extrapolation.κ=∞is reached in the level crossing point LCP={a k(κ)≅18.226,q k(κ)≅0.9986}which separates plateau-regime LCP≤a≤1/T from the∞-RSB continuum0≤a≤LCP.The Parisi box size m(T=0.03)=0.5467is the break point value.Inserts illustrate that q l-and a l-levels meet in the LCP forκ=∞.for the subclass of large parameters a l(their number also grows to infinity asκ→∞).The linear temperature decay of all Parisi box sizes m l(κ,T)∼T holds for allfinite RSB-orders,but not all m′s should vanish in the RSB limit at zero temperature, since the break point is not expected to be at m1=0 (even in the T→0-limit50).Thus,one should describe a non-commuting limits T→0andκ→∞properly. The Taylor series,valid as a low temperature expansion for anyfixedfinite RSB-order,m l(κ,T)≡a l(κ,T)T=a l(κ,0)T+11+d1,l x+d2,l x2.(17) This formfits well the availablefinite T data up to 50-RSB and satisfiesf al(0)=c0,lfinite and f al(x)∼1subsectionV A that all ratios become level-independent in the large x limitm l(x)a l−1(x)=f al(x)κf i H2/3112κ l =1a l (q 2l −q 2l +1)(24)⇒lim κ→∞U (κ,0)=−12∞da (1−q 2(a ))−M (H )H,(26)where M (H )denotes the field-generated magnetization.Recalling the small-a expansion of the order function,q (a )∼a −const.a 3+O (a 5),one must expect an H 10/3-contribution from the plateau-regime,which implies also an O (κ−5)contribution.The free energy data are com-patible with an H 10/3small field scaling part.It must be concluded that the leading correction κ−4must originate in the intermediate a -regime (not yet identified in detail).It can,after all what was said before,not be assumed to be a scaling contribution.We should therefore attribute it to the regular free energy part.The entropy was found to obey 30S (κ,T =H =0)=−1(κ+κ0)4+c 5。

相关文档
最新文档