Announcements

Syllabus

Class policies

Schedule

Resources

Course Schedule

B553 Spring 2011

(subject to change)

 

No.

Date

 

Subject

Out

In

Readings

1

10-Jan

Tu

Introduction, mathematical preliminaries

 

 

Notes 1

2

12-Jan

Th

Univariate optimization

HW1

 

Notes 2

3

17-Jan

Tu

Root finding

 

 

 

4

19-Jan

Th

Mathematical preliminaries, pt 2

 

 

Notes 3

5

24-Jan

Tu

Gradient descent

HW2

HW1

Notes 4

6

26-Jan

Th

Second-order optimality conditions

 

 

Notes 5

7

31-Jan

Tu

Newton and quasi-Newton methods

 

 

Notes 6

8

2-Feb

Th

Constrained optimization, Lagrange multipliers, KKT conditions

HW3

HW2

Notes 7

9

7-Feb

Tu

Linear programming, quadratic programming, sequential quadratic programming

 

 

Notes 8

10

9-Feb

Th

Convex programming, interior point methods

 

 

Notes 9

11

14-Feb

Tu

Simulated annealing, genetic algorithms

HW4

HW3

Notes 10

12

16-Feb

Th

Optimization of stochastic functions

 

 

Notes 11

13

21-Feb

Tu

Global optimization, branch and bound techniques

 

 

Notes 12

14

23-Feb

Th

Review of probability

 

 

K&F 2

15

28-Feb

Tu

Bayesian networks

HW5

HW4

K&F 3.1-3

16

1-Mar

Th

Representations and exact inference

 

 

K&F 5.1-3

17

6-Mar

Tu

Variable elimination and clique trees

 

 

K&F 9.1-4

18

8-Mar

Th

Inference via message passing

 

HW5

K&F 10.1-3

---

13-Mar

Tu

 

 

 

 

---

15-Mar

Th

 

 

 

 

19

20-Mar

Tu

Monte Carlo methods

HW6

 

K&F 12.1-3

20

22-Mar

Th

Continuous/hybrid distributions

 

 

K&F 7,14.1-2

21

27-Mar

Tu

 

 

 

 

22

29-Mar

Th

Parameter learning

HW7

HW6

K&F 17.1-4

23

3-Apr

Tu

Bayes net structure learning

 

 

K&F 18.1-4

24

5-Apr

Th

Expectation maximization algorithm

 

 

K&F 19.1-2

25

10-Apr

Tu

Markov chains/HMMs/DBNs

HW8

HW7

K&F 6.2,15.1

26

12-Apr

Th

Viterbi algorithm, Kalman filtering, particle filtering

 

 

K&F 13.2, 15.3-4

27

17-Apr

Tu

Special topics lectures

 

 

28

19-Apr

Th

Recap/ Special topics lectures

 

 

 

29

24-Apr

Tu

Final project presentations

 

HW8

 

30

26-Apr

Th

Final project presentations

 

 

Homework assignments

HW1. Math review and univariate optimization

HW2. Unconstrained optimization on a nonlinear least-squares problem

HW3. Lagrange multipliers and linear programming.

HW4. Empirical comparison of stochastic optimization techniques for neural network training.

HW5. Probabilistic inference with Bayesian networks

HW6. Monte Carlo methods for computing a Bayesian posterior

HW7. Bayes net parameter and structure learning

HW8. Hidden Markov models for anomalous event detection.