機器學習(英文授課)

Machine Learning

學期

107-1

學分

0 學分

當期課號

5229

永久課號

IOC5191

開課單位

資訊科學與工程研究所

授課教師

洪瑞鴻、邱維辰

校區

光復

類別

選修

上課時間表

節	週一	週四
3 10:10–11:00		機器學習(英文授課) EC115（光復） 2 節連堂
4 11:10–12:00		機器學習(英文授課) EC115（光復） 2 節連堂
7 15:30–16:20	機器學習(英文授課) EC115（光復）

* 根據陽明交大上課時間表所列

概述

(1) To build big picture on machine learning field and equip with the ability of implementation machine learning techniques. This course will introduce the theory behind the techniques, so a great deal of time will spend on the mathematics foundation. (2) To understand the properties of different learning algorithms and learn how to use, when to use, which to use, under different scenarios.

先修科目

Calculus, Probability, Statistics, Linear Algebra, Introduction of Machine learning or equivelent

備註

無備註

教學方式

教師未提供此項資料

評分方式

Involvement (10%), homework (30%) mid term (30%), final (30%)

課程大綱

Probability and information theory

Basics of Machine learning, Shannon Entropy, Bayes' theorem, Naive Bayes classifier

講授：

6
Regression and classification

Maximum likelihood, SSE, Linear regression, BIC, Logistic regression, overfitting / regularization (Ridge and Lasso), model selection

講授：

6
Dimension reduction and feature extraction

SVD, FFT, PCA and NMF

講授：

6
Distribution and Statistics

Gaussian integral, Central limit theory, Gaussian distribution, Moments of distribution, moment generation function

講授：

6
Kernel methods

Kernel methods, dual representation, Support vector machine, support vector regression

講授：

6
Generative Models

understand the data from modelling formation, GMM, EM, graphical model, HMM, topic model

講授：

6
Inference

sampling methods, MCMC, Gibbs sampling, relation to clustering

講授：

6
Deep Learning

NN, DNN, CNN, RNN, VAE, GAN

講授：

6

週次計畫

週次	主題
第 1 週	1. My teaching method, overview of the course 2. Basics of probability (Joint, conditional probability and independence) 3. Basics of information theory (Entropy, relative entropy, mutual information) 4. Bayes theorem (Maximum likelihood, conditional independence, naive Bayes classifiers, Bayesian network) 9/11 9/14 9/18 9/21
第 3 週	1. Classification and Regression 2. Linear regression (MLE) 3. Logistic regression 4. Regularization (MAP, Ridge and Lasso) 5 .Basics of optimization 9/25 9/28 10/2 10/5
第 5 週	1. Correlation Coefficient 2. PCA 3. NMF 4. FFT 10/9 10/12 10/16 10/19
第 7 週	1. Distribution (Beta, Gaussian, etc.) 2. Moment Generation Function 3. Special function 4. Conjugate prior 10/23 10/26 10/30 11/2
第 9 週	Midterm 11/6 11/9
第 10 週	1. Kernel Method 2. Support Vector Machine 3. From Binary to Multi-class Classification Based on SVM 4. Support Vector Regression 11/13 11/16 11/20 11/23
第 12 週	1. Introduction to Generative Model 2. Gaussian Mixture Model 3. Expectation and Maximization 4. Graphical Models 5. Hidden Markov Model 6. Topic Models (Nonparametric Bayesian) 11/27 11/30 12/4
第 14 週	1. Posterior Inference for Generative Model 2. Relation to Clustering And Dimension Reduction 3. Introduction to Monte Carlo methods 4. Markov Chain Monte Carlo 5. Gibbs Sampling 6. Simulated Annealing 12/11 12/14 12/18
第 16 週	1. From Machine Learning to Deep Learning 2. Neural Networks 3. Convolutional Neural Networks 4. (Stochastic) Gradient Descent 5. Brief Introduction to Sequential-to-Sequence Models 5. Brief Introduction to Deep Generative Models 12/25 12/28 1/1 1
第 18 週	Final exam 1/8 1/11

教科書

[1] Christopher Bishop, Pattern Recognition and Machine Learning, Springer, 2007 [2] A. Smola and S.V.N. Vishwanathan, Introduction to Machine Learning, Cambridge University Press, Oct. 2010

Office Hours

地點: office
時間: TBA
聯絡方式: email