Courses: SYSC 610: Discrete Multivariate Modeling-II -- Spring 2009
From description of SySc 651 (which is DMM-I): In this course, information theory is used as a framework for modeling and data mining: for analyzing static or dynamic relations among discrete variables, for detecting complex interaction effects, and for discovering nonlinearities in continuous variables made discrete by binning. In the systems literature, these information-theoretic and related set-theoretic methods, used together with graph theory techniques, are called “Reconstructability Analysis” (RA).
SySc 610 continues the presentation of discrete multivariate modeling (SySc 551/651), and will focus on (a) student projects and (b) lectures on advanced topics.
Projects will aim at conference or journal publications. Possible projects are:
- intensive analysis of some dataset
- software development that enhances the current set of RA tools
- performance study of an RA method (e.g., Fourier RA, BDD-SRA)
- theoretical study on RA methods
Lectures will include most if not all of the following advanced topics (& perhaps others):
- more on state-based RA & SB-RA software; SB-RA & decision analysis
- continuous functions (DVs): expected value, k-systems, & j-systems approaches
- set-theoretic RA: binary decision diagrams; 'modified' SRA; approximating IRA
- Fourier-based (regression-like) RA calculations
- more on loopless models: algebraic techniques & 'dependency analysis'
- intra-model analyses (Krippendorff)
- modeling with latent variables
- RA used with genetic algorithms & neural networks
- RA analysis of cellular automata dynamics
- illustration of RA time-series analysis
- OCCAM software architecture
- identification with inconsistent data
Prerequisites: SySc 551/651 (or – with permission of the instructor –solid background in log-linear modeling, Bayesian networks, or related methods)
TEXT (aside from SySc 551/651 texts): Materials not on web will be distributed in class.