Computer Science 246/446

Machine Learning

Spring 2014

Instructor: Dan Gildea office hours Mon 2-3:15pm
TA: Yuancheng Li, office hours Mon 2-3pm, Wed 2-3pm, CSB 617
Location: M/W 3:25pm-4:40pm, Dewey 2210E

Homeworks

  • Python help session: 3:25pm Fri 1/17, CSB 601

Required text: Christopher M. Bishop, Pattern Recognition and Machine Learning.

The following are useful references in addition to the reading material assigned for each class:

  • Stuart Russell and Peter Norvig, Artificial Intelligence, A Modern Approach.
  • Trevor Hastie, Robert Tibshirani, Jerome Friedman, The Elements of Statistical Learning: Data Mining, Inference, and Prediction.
  • Larry Wasserman, All of Statistics, 2004.
Readings not taken from Bishop are available through nb.

Syllabus

Onwe will coverwhich means that after class you will understandif before class you have read
1/15 Probability Theory independence, bayes rule bishop 1.2, 1.4, app E
1/22 Information Theory entropy, kl-distance, coding bishop 1.6
1/27 Gaussians covariance, conditioning bishop 2.3
1/29 Probabilistic Inference priors: bayesian reasoning, MAP bishop 3.4
2/3 Perceptron stochastic gradient descent bishop 3.1, 4.1
2/4 Backpropagation DP for gradient descent bishop 5.1, 5.2, 5.3
2/10 Support Vectors strong duality bishop 7.1
2/12 Support Vectors the kernel trick bishop 6.1, 6.2
2/17 Probabilistic Inference message passing bishop 8.4
2/19 Tree decomposition cyclic graphs koller and friedman
2/24 Expectation Maximization L = Q + H + D bishop 9
2/26 Expectation Maximization mixture of gaussians bishop 9
3/3 Sampling Markov Chain Monte Carlo bishop 11.2
3/5 Sampling Gibbs Sampling bishop 11.3
3/17 Sampling cont
3/19 Review
3/24 Midterm
3/26 Midterm Solutions
3/31 Learning Theory VC dimension Kearns and Vazirani
4/2 Logistic Regression maximum entropy bishop 4.3
4/7 Hidden Markov Models forward-backward bishop 13.2
4/9 HMM cont'd minimum bayes risk
4/14 Optimization Newton's method, DFP Nocedal 8.1, 9.1
4/16 Optimization cont'd LBGFS
4/21 Particle Filters slam Thrun et al. ch 4
4/23 PCA Eigenvectors
4/28 Reinforcement Learning q-learning sutton ch 3, 4.3, 4.4, 6.1, 6.5, 7.2, 11.1
4/30 Review come to class with questions!
Final Exam: Monday May 5, 8:30-11:30am, Dewey 2210E.

Grading

  • Final exam: 30%
  • Homeworks: 45%
  • Midterm: 20%
  • Class participation: 5%

gildea @ cs rochester edu
April 23, 2014