“Estimation is art; detection is science”
Graduate level course in statistical signal processing. Focusses on detection and estimation theory, and the relationships between them. Concentration on discrete-time results. Performance bounds derived from signal processing and information theoretic perspectives.
Prerequisites: Knowledge of random processes.
Meets TTh 1-2:15PM, Keck 101
- Random processes, in both continuous- and discrete-time
- Second-order description: expected value, correlation function and power density spectrum
- Sampling processes
- Jointly defined processes: cross-correlation function, linear filtering
- Karhunen-Loève expansion
- Constrained and unconstrained optimization
- Estimation theory
- Parameter estimation: mean-squared error, MAP, maximum likelihood
- Linear estimation
- Spectral estimation
- Adaptive filters
- Optimal filtering: Wiener, Kalman and Bayesian filters; denoising
- Compressive sensing
- Detection theory
- Statistical hypothesis testing
- Sequential detection
- Detection of signals in noise
- Unknown parameters
D.H. Johnson. Statistical Signal Processing. Available as a downloadable pdf file.
Lecture videos can be accessed here. Note that they are sorted by date, with the most recent being listed first.
We use piazza as the course’s online discussion forum. Questions that arise outside of class can be posted there. Students as well as instructors can provide answers that all can see.
Grades are based on two take-home quizzes, the problem sets (usually weekly), and an oral final. Problem set grades will be separately averaged and normalized to a 100 maximum, the grading standard for the quizzes and the final.
University Disability Accommodation Policy
Any student with a documented disability needing academic adjustments or accommodations is requested to speak with the course instructor during the first two weeks of class. All discussions will remain confidential. Students with disabilities should also contact Disabled Student Services in the Ley Student Center.