September 1
Questions?
Introduction to LaTeX.
Joint probability, conditional probability, Bayes rule.
|
| Bishop, Chapter 1. (skip Section 1.6 for now)
September 3
|
Linear models for prediction.
|
| Bishop, Sections 3.1-3.2.
Assignment 1
due by midnight for on-campus students. Submit via RamCT.
September 8
Linear models continued.
|
|
Assignment
1 due by midnight for distance-learning students. Submit via RamCT.
September 10
R Tips for Linear Models
Backing up files
Collinearity
Assignment 2 questions
|
|
September 15
Questions on Assignment 2.
Nonlinear inputs for linear models
Probabilistic
framework for linear models
|
|
Assignment
2 due by midnight for on-campus students. Submit via RamCT.
September 17
Classification with linear least squares,
quadratic
discriminant analysis,
linear discriminant analysis.
|
| Bishop 4.1-4.3
Assignment
2 due by midnight for distance-learning students. Submit via RamCT.
September 22
Questions on Assignment 3.
Bayesian Regression
|
| Bishop 2.3.3, 3.3.1, 3.3.2
September 24
Questions on Assignment 3.
Logistic Regression for Classification
|
| Bishop 4.3.2
September 29
Questions on Assignment 3
More Logistic Regression
|
| Bishop 4.3.4
Assignment
3 due by midnight for on-campus students. Submit via RamCT.
October 1
| Scaled Conjugate Gradient
|
|
| Assignment
3 due by midnight for distance-learning students. Submit via RamCT.
October 6
k-Nearest-Neighbors algorithm and R implementation
Rgtk
|
| Bishop 2.5.2
October 8
| Neural Networks for Nonlinear Regression
|
| Bishop 5.1-5.3
October 13
| Neural Networks
|
|
Assignment
4 due by midnight for on-campus students. Submit via RamCT.
October 15
Neural Networks, Overfitting.
|
| Bishop 5.5.1, 5.5.2
Assignment
4 due by midnight for distance-learning students. Submit via RamCT.
October 20
Reducing the dimensionality of data
|
| Bishop 12.1, 4.1.4-4.1.6
October 22
PCA and Fisher
|
|
October 27
Neural Networks for Nonlinear Classification
|
|
Assignment
5 due by midnight for on-campus students. Submit via RamCT.
October 29
Nonlinear Logistic Regression with Neural Networks
|
|
Assignment
5 due by midnight for distance-learning students. Submit via RamCT.
November 3
Introduction to reinforcement learning
|
|
November 5
Representing value functions as tables
|
| Reinforcement
Learning: An Introduction, by Sutton and Barto, Chapters 1 and 2.
You can also obtain a higher-quality pdf version of this book through
Morgan library. Visit this
page and click on the "View electronic book" link.
November 10
Representing value functions with neural networks
|
| Sutton and Barto, Chapters 3, 4, 6
Assignment
6 due by midnight for on-campus students. Submit via RamCT.
November 12
Neural Q gradient. Simple maze examples.
|
| Sutton and Barto, Chapters 5, 7, 8
Assignment
6 due by midnight for distance-learning students. Submit via RamCT.
November 17
New Assignment. Discussion of R code for solving marble
control problem.
|
|
November 19
Assignment 7 questions.
Introduction to Genetic Algorithms
|
|
November 23-27
Thanksgiving Break
December 1
Questions on Assignment 7. Barreto's Restricted Gradient
Descent algorithm for learning value functions.
|
| Restricted
Gradient-Descent Algorithm for Value-Function Approximation in
Reinforcement Learning by Barreto and Anderson, Artificial
Intelligence, volume 172, numbers 4-5, March 2008, pp. 454-482
December 3
Distributed computing in R with Snowfall
|
|
December 8
Questions on Assignments 7 and 8. More on Snowfall.
Support vector machines.
|
| Bishop 7.1, 6.1, 6.2
Assignment
7 due by midnight for on-campus students. Submit via RamCT.
| December 10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 
