Back to Courses


CDS 270-3

CDS 270-3 -- Spring 2007
Third Term

Stochastic System Analysis and Bayesian Updating


Course Description

CDS 270-3. Stochastic System Analysis and Bayesian Updating. Units 3-0-6.

Course Outline

This is a new course that focuses on a probabilistic treatment of uncertainty in modeling a system's behavior and its excitation. We will start by examining the foundations of probability as a multi-valued logic for plausible reasoning with incomplete information; this will lead to a rigorous meaning for the probability of a model for a system. We will then focus on new approximate analytical and stochastic simulation tools for robust system analysis and Bayesian system identification that have been developed over the last decade or two, primarily by statisticians, computer scientists and engineers. The topics covered will include: Bayesian updating of models of a system to predict its future behavior based on measured response, including new Markov Chain stochastic simulation techniques; Bayesian model class selection with a new information-theoretic interpretation that shows that it automatically gives a quantitative Ockham principle of model parsimony; recently-developed stochastic simulation techniques for evaluating the response of stochastic dynamic systems subject to stochastic excitations, especially computing small failure probabilities (i.e. probability of the system reaching some undesirable state); and Bayesian sequential estimation of system states and model parameters, generalizing the Kalman filter.

Tuesday, Thursday 9 AM - 10:30 AM; 306 Thomas
James L. Beck

Office hours: by appointment
Teaching Assistants
Alex Taflanidis, Sai-Hung (Joseph) Cheung

Office hours: by appointment


  • The first lecture will be on Tuesday, March 27, 2007 in 306 Thomas at 9 AM.

Course Material