ELE 530: Theory of Detection and Estimation
Course Description
In this course we investigate how to use the tools of probability and signal processing to estimate signals and parameters and detect events from data. In many cases we can identify the optimal estimator/detector or at least bound the performance of any estimator/detector.
Instructor
Prof. Paul Cuff
Office location: B-316 E-quad
Office hours: M/W 4:30-5:30pm (after the lectures).
Time and Location
Lectures
J-201 (in the J-wing of the E-quad)
M/W 3-4:20pm
Text
We will be covering topics found in a variety of sources. If you do not wish to buy the entire library of books listed below, the main text (Kay Vol. 1) will be a useful one to have on hand.
Main text
Fundamentals of Statistical Signal Processing - Estimation Theory (Vol. 1) Steven M. Kay, Prentice Hall, 1993.
 |
Other text
Fundamentals of Statistical Signal Processing - Detection Theory (Vol. 2) Steven M. Kay, Prentice Hall, 1998.
Linear Estimation Kailath, Sayed, and Hassibi, Prentice Hall, 2000.
An Introduction to Signal Detection and Estimation H. Vincent Poor, 2nd Edition, Springer, 1998.
Elements of Information Theory Cover and Thomas, 2nd Edition, Wiley, 2006.
EE 278 lecture notes available in Blackboard, Abbas El Gamal, Stanford.
Blackboard
Blackboard will only be used for communicating grades on assignments and exams and for distributing solutions (not intended for the eyes of future students).
Topics
Bayesian
Minimum mean square-error (MMSE)
Linear MMSE
Hilbert space of random variables
Minimum probability of error (MAP)
Signals
Stationarity and power spectral density
Wiener filter
Kalman filter
Non-Bayesian Estimation
Sufficient Statistic
Bias
Minimum variance unbiased estimator
Cramer-Rao bound
Maximum likelihood
Efficient estimator
Non-Bayesian Detection
Hypothesis test
Neyman-Pearson lemma
Likelihood ratio test
Kullback-Leibler divergence
Matched filter
Sequential test
Methods of dealing with complexity
Expectation maximization
Hidden Markov model
Graphical models