Supervised Learning

Instructor: Dr. Jianlin Cheng

Location:Lafferre Hall E3403, Time: MWF: 11 am - 11:50 am, Office Hours: by appointment, Semester: Fall 2019

Syllabus

Lecture Slides

Acknolwledgments: the slides are customized from the materials of CMU machine learning courses taught by Drs. Singh, Xing and Mitchell, MU machine learning courses taught by Cheng, Cheng's research, the text book, and Stanford machine learning courses.

1. Introduction and Bayes optimal learning

2. Learning distribution, parametric learning, maximum likelihhod estimation (MLE) and maximum a posterior (MAP), Naive Bayes Classifier, and Generative Classifier (textbook Chapters 1,2)

3. Discriminative Classifier and Logistic Regression (textbook Chapter 4)

4. Linear and Non-Linear Regression

5. Non-Parametric Methods for Density Estimation, Classification and Regression

6. Model Selection (PDF) and [PPT version]

7. Boosting (PDF) and [PPT version]

8. Deep Learning Networks (PDF) and [PPT version]

9. Support Vector Machine I and [PPT version]

10. Support Vector Machine II

11. Hidden Markov Models

12. Bayesian Networks and Graphical Models

Assignments

All the assignments should be submitted to mumachinelearning@gmail.com.
  • Assignment 1 (due Sept. 1, 2019)
  • Assignment 2 ( due Sept. 11)
  • Assignment 3 (see Slides of Lecture 4, due Sept. 28)
  • Assignment 4 (write a description of your course project (see the document for the requirements), due Oct. 18)
  • Project

  • A popular data source: UCI Machine Learning Data Repository
  • General instructions: Apply or develop at least one machine learning method to a classification or regression problem. You may work on a public data set (e.g. one in the UCI Machine Learning Data Repository) or your own research project. You may form a group of up to 4 students to work on the project.
  • Presentation and Report

    Each group / person has 10 minutes to present the selected project. The last several classes will be reserved for presentation. The final project report will be due on Dec. 8. A report should include a title, author names, an abstract, an introduction, method description, results, conclusion, and references if any.

    Machine Learning Background

    A Set of Statistical Machine Learning Tutorials (taught by Andrew Moore at CMU and Google).

    Introduction to probability distribution (Chapter 2).

    Machine Learning for Bioinformatics.

    Optmization Methods for Machine Learning and Deep Learning.

    Related Courses taught by Prof. Jianlin Cheng

  • Supervised Machine Learning
  • Computational Modeling of Molecular Structures
  • Data Mining and Knowledge Discovery
  • Machine Learning for Bioinformatics
  • Problem Solving in Bioinformatics
  • Computational Optimization Mehtods