Machine Learning & Data Science

Instructor: Guan-Ju Peng (彭冠舉) @ Applied Math Department, NCHU

email: gjpeng@email.nchu.edu.tw

Textbook:

1. Python Machine Learning, Sebastian Raschka

     2. Introduction to Machine Learning, Ethem Alpaydm.

• Introduction to Machine Learning 

• Fundamental Learning Theory

• Linear Model 

• No-Free-Lunch Theorem

• Convex Learning Problems

• Support Vector Machine and Decision Tree

• Data Preprocessing

• Dimension Reduction

• Tuning

• Ensemble Learning

• Cluster Analysis 

• Regression

上課連結: 週六班

PART I

Week 1: 

       2/18補行上班日

Week 2:  

       Course Logits

       What is Machine Learning?

       Type of ML 

       (自主學習) Implementation: Data Plotting

       Formal Model 

Week 3:

       Perceptron (I II III)

       (自主學習) Implementation: Perceptron

       Adaline and Gradient Descent

       (自主學習) Implementation : Adaline

       (自主學習) Implementation : Adaline SGD

Week 4:

      Realizability Assumption 

      Sample Complexity 

      Probably Approximately Correct (PAC) Learning

Week 5:      

      Agonostic PAC Learning 

      Uniform Convergence ( 1 2 )

Week 6: 

      3/25補行上班日

Week 7: 

      4/1 因應連續假期放假本課程停課一次 

      請各位同學可以自行找時間觀看下列三個影片

      General Linear Model  

      Logistic Regression 

  o Implementation: A General Flow of Learning Process

 作業 I:

     Midterm I (期末最後一次上課繳交)

PART II

Week 8:  

     No Free Lunch Theorem

     Bias-Variance TradeOff    

Week 9: 

     Convex Learning Problems

     Surrogate Function 

Week 10: 

     Stability of Strongly Convex Learning Problems 

     Support Vector Machine

     Kernel Trick

Week 11: 

     The Representer Theorem

     KNN

     Decision Tree

 作業 II:

     Midterm II    

Week 12:  

     Data Preprocessing

     L1 Regularization 

     (自主學習) Implementation: Feature Selection

     Principle Component Analysis

Week 13: 

     LDA

     Kernel PCA

     (自主學習) Implementation: Kernel PCA

     Machine Learning Pipeline and Cross-Validation 

     Learning Curve

     Imbalanced Data

Week 14: 

     Majority Voting

     Bagging      

     AdaBoost

     Clustering and K-means

     Determine the number of clusters

     (自主學習) Implementation: Majority Voting and Parameter Tuning 

Week 15:     

     Agglomerative Clustering

     DBSCAN     

     Regression ( I II )

     Multi-Layer Perceptron 

Week 16:     

     Non-Linearity of MLP 

     Implementation: Retrieving Data 

     Forward Propagation and Cost

     Back Propagation ( I II ) 

     Cost and Objective     

Week 17: Final Report 

作業 III 

      Final Exam