Introduction. Overview of learning (ch.
1)
lecture slides: 1
per page or 4
per page |
Concept learning, version spaces (ch.
2)
lecture slides: 1
per page or 4
per page
|
| Version Spaces, inductive bias (ch. 2). |
Tutorial in Information Theory.
lecture slides: 1
per page or 4
per page |
| Tutorial in Information Theory (cont.) |
Decision trees, overfitting, Occam's razor
(ch. 3).
lecture slides: 1
per page or 4
per page |
Decision trees, overfitting, Occam's razor
(cont.).
|
Neural networks (ch. 4)
lecture slides: 1
per page or 4
per page |
| Neural networks cont. |
| Neural networks cont. |
| NO CLASS |
PAC learning (ch.7 through 7.3)
lecture slides: 1
per page or 4
per page |
PAC learning, VC dimension, Mistake
bounds.
(ch. 7.4 through 7.4.3, 7.5 through 7.5.3) |
| NO CLASS |
| Review for mid-term |
| Mid-term |
Statistical Estimation and Testing (ch.
5).
lecture slides: 1
per page or 4
per page |
Bayesian learning: MAP and ML learners (ch.
6)
lecture slides: 1
per page or 4
per page |
| Bayes learning examples; MDL (ch. 6) |
| Bayes Optimal Classifier, Gibbs sampling. |
| Naive Bayes and learning over text (ch
6) |
| Hidden Markov Models (HMM) (ps,pdf) |
HMM (cont.); Examples from Speech Recognition
|
Instance based learning, k nearest nbr., locally
weighted regression,
radial basis functions (ch. 8)
1 per
page or 4 per
page.
|
| E-M algorithm (ch 6) |
| NO CLASS |
Reinforcement learning (ch. 13)
1 per page or 4
per page |
Combining
Learned Classifiers, Weighted
Majority, Bagging,
Boosting (1),
(2)
(ch. 7: Weighted majority)
|
Genetic algorithms, genetic programming (ch.
9)
lecture slides: 1
per page or 4
per page |
| Final review |
| Final exam (1-4pm). |
| <=========NOTE: the yellow highlighted arrows must match,
or else your browser may be mis-aligning dates to lectures! |