6.972 Algorithms for Estimation and Inference (H)

L TR1-2:30, Room 56-154, * Rec W11, 1 and 2, Room 12-102
Professors P. Golland, A. S. Willsky and G. W. Wornell
Prereq.: 6.011 and 18.06
4-0-8

This subject qualifies as a Communication, Control and Signal Processing concentration subject.

Estimation and inference problems arising in signal processing, optimization and control, and machine learning. Second-order characterizations of random phenomena. Least squares estimation: Orthogonality, and whitening; Wiener filtering; estimation for state space models: Kalman filters and smoothers, properties, and efficient algorithms. Model estimation: ergodicity, spectral estimation, likelihood calculation, all-pole models and the Levinson algorithm. Estimation for Markov models: particle filters, Viterbi algorithm. Markov random fields and graphical models: Belief Propagation Algorithms and properties; exponential families, variational methods, and max-entropy modeling.

*Note: the first recitations will be on Friday, September 8 (times/locations TBA). All subsequent recitations are on Wednesdays at the times and locations listed above. This course can be used for TQE credit for EE and for CS.