使用+HTS+进行中文语音合成之研究

合集下载
  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)。


Concepts
視其所以,觀其所由,察其所安,人焉廋哉?人焉廋哉? 近朱者赤、近墨者黑

Artificial intelligence Machine learning
2/10
More about Machine Learning

Quick examples
Use height and weight to predict gender Use historical data to predict stock market



Model construction

Classifier

Structure id
Feature selection Feature extraction

Missing value imputation


Visualization
Outlier removal Feature normalization


Comparison


Before ML Rule-based After ML Data-driven Probability & statistics Linear algebra Optimization …
3/10

Prerequisites for ML

Broad Areas of ML
y*

Two steps in modeling
structure identification: input selection, model complexity parameter identification: optimal parameters

5/10
Example of Modeling


For researchers
Design new learning algorithms Speed up the learning process Derive theoretical bound of accuracy Take care of data

Big data Imbalanced data Missing data …

Info. retrieval
Recommendation search


Security
Fraud detection (credit card, telephone) Network intrusion Video surveillance Speaker id.

8/10
First Look at a Dataset

GHW (gender-height-weight) dataset

10000 entries with 3 columns
Gender: Categorical Height: Numerical Weight: Numerical
Source: https://www.slideshare.net/tw_dsconf/ss-70083878
6/10
More ML Application Examples
Optical character recognition Document classification Part-of-speech tagging Speech processing

4/10
Concept of Modeling

Given desired i/o pairs (training set) of the form (x1, ..., xn; y), construct a model to match the i/o pairs
x1
... xn y
Unknown target system Model f()

Parameter id

Cross validation for perf. evaluation
Confusion matrix DET/ROC plots PR plot

Error analysis
12/10
One Step Further AutoML

AutoML

Machine learning for automated algorithm design

Regression: Predict a real-value for a given object


Ranking: Order objects according to some criterion

Clustering: Partition data into homogeneous groups Dimensionality reduction: Find low-dimensional manifold preserving some properties of the data Density estimation: Learning probability density function according to sample data

DNN

Radial-basis function networks Random forests
11/10

Classification and regression trees (CART)


A Canonical Way to ML

Data processing
Dataset collection Feature extraction

Keys
Transfer learning Learning to learn Hyper-parameter tuning


Resources

ML4AAD Google’s Cloud AutoML
Official site Media coverage

13/10
An Example & On-going Projects
K-nearest-neighbor classifiers Quadratic classifiers


Naï ve Bayes classifiers

Linear classifiers
Single-layer perceptrons SVM


Neural networks

Multilayer perceptrons


Source
9/10
Multiple Ways of Prediction

Multiple ways of prediction based on GHW dataset
Classification: Height, weight gender Regression: Gender, height weight Regression: Gender, weight height
wenku.baidu.com
Classification: Assign a category to a given object

Determine if it will rain tomorrow Predict the exchange rate Rank the webpages returned by a search engine

Definition
Field of study that gives computers the ability to learn without being explicitly programmed – by Arthur Samuel, 1959. Computational methods that use existing data to make predictions

Recognition synthesis
Face recognition

Image recognition

Self-driving cars
7/10
What People Do with ML

For practitioners
Acquire application domain knowledge Know properties of machine learning methods Get familiar with ML tools
Intro to Machine Learning
J.-S. Roger Jang (張智星) jang@mirlab.org http://mirlab.org/jang
MIR Lab, CSIE Dept.
National Taiwan University
Machine Learning (ML)

An example of canonical ML (with a flavor of AutoML)

Leaf identification

On-going projects
Customer’s behavior prediction (Esun Bank) Plant image recognition (ITRI) Speaker verification (China Trust) Healthcare and medical data analytics (NTU Hospital)


In general
Classification: the output is categorical Regression: the output is numerical


Sometimes it’s not so obvious!
10/10
List of Classifiers

Commonly used classifiers

14/10
相关文档
最新文档