Lectures
- 1. Overview
- 2. Linear Regression, Gradient Descent, Normal Equations
- 3. Locally Weighted Regression, Linear Regression, Logistic Regression
- 4. Newton's Method, Exponential Family, General Linear Models
- 5. Generative Algorithms, Gaussian Discriminant Analysis
- 6. Applications of Neural Network, Support Vector Machine
- 7. Optimal Margin Classifier, Karush-Kuhn-Tucker (KKT) Conditions, SVM Dual
- 8. Non-linear Decision Boundaries and Soft Margin SVM
- 9. Bias-variance Tradeoff, Empirical Risk Minimization, The Union Bound, Hoeffding Inequality
- 10. Uniform Convergence, VC Dimension, Model Selection
- 11. Bayesian Statistics and Regularization, Applying Machine Learning Algorithms
- 12. Unsupervised Learning, Mixtures of Gaussians and the EM Algorithm, Jensen's Inequality
- 13. Mixture of Gaussian, Mixture of Naive Bayes, Factor Analysis Model
- 14. Factor Analysis Model, EM for Factor Analysis, Principal Component Analysis
- 15. Independent Component Analysis (ICA)
- 16. Reinforcement Learning, Markov Decision Process, Value Function
- 17. Continuous States, Discretization & Curse of Dimensionality
- 18. State-action Rewards, Finite Horizon MDPs, Dynamical Systems
- 19. Debugging Reinforcement Learning (RL) Algorithm, Linear Quadratic Regularization (LQR),
- 20. Partially Observable MDPs, Reinforce Algorithm, Pegasus Algorithm
Machine Learning
Course Summary
This course is based on CS229 Machine Learning made available by Stanford University: Stanford Engineering Everywhere under the Creative Commons Attribution 3.0 license.
This course in Artificial Intelligence provides a broad introduction to machine learning and statistical pattern recognition and will cover recent applications of machine learning, such as to robotic control, data mining, autonomous navigation, bioinformatics, speech recognition, and text and web data processing. Students are expected to have the following background: Knowledge of basic computer science principles and skills, at a level sufficient to write a reasonably non-trivial computer program; Familiarity with the basic probability theory; Familiarity with the basic linear algebra.
Topics include: supervised learning (generative/discriminative learning, parametric/non-parametric learning, neural networks, support vector machines); unsupervised learning (clustering, dimensionality reduction, kernel methods); learning theory (bias/variance tradeoffs; VC theory; large margins); reinforcement learning and adaptive control.
Reading Material
Not available.Course Material
1. Linear Regression, Classification and logistic regression, Generalized Linear Models2. Generative Learning algorithms
3. Support Vector Machines
4. Learning Theory
5. Regularization and model selection
6. The perceptron and large margin classifiers
7. The k-means clustering algorithm
8. Mixtures of Gaussians and the EM algorithm
9. The EM algorithm
10. Factor analysis
11. Principal components analysis
12. Independent Components Analysis
13. Reinforcement Learning and Control
14. Review - Linear Algebra Review and Reference
15. Probability Theory Review
16. Matlab Review
17. Convex Optimization Overview, Part I
18. Convex Optimization Overview, Part II
19. Review - Hidden Markov Models
20. Review - Gaussian Processes
21. Assignments
Assignment | Assignment Data Files | Solution | Solution Data Files |
Problem Set 1 | PS1-data.zip | Solution Set 1 | ps1_solution-data.zip |
Problem Set 2 | PS2-data.zip | Solution Set 2 | |
Problem Set 3 | PS3-data.zip | Solution Set 3 | ps3_solution-data.zip |
Problem Set 4 | PS4-data.zip | Solution Set 4 | ps4_solution-data.zip |
Other Resources
1. Advice on applying machine learning (313 KB pdf)Slides from Andrew's lecture on getting machine learning algorithms to work in practice can be found here.
2. Previous projects
A list of last year's final projects can be found here.
Software
1. Matlab Tutorial2. A Practical Introduction to Matlab
3. GNU Octave
GNU Octave is a high-level language, primarily intended for numerical computations. It provides a convenient command line interface for solving linear and nonlinear problems numerically, and for performing other numerical experiments using a language that is mostly compatible with Matlab.
4. UCI Machine Learning Repository
The UCI Machine Learning Repository contains a large collection of standard datasets for testing learning algorithms.
Discussion Forum
For discussion on this topic, please go to the relevant forum for Machine Learning. Click the button below to open the forum page in a new window.