Research Publications

This report is the background theory for Discrete Component Analysis software called DCA. Currently the software is run in stand-alone mode, and scavengers data streaming libraries and Dirichlet utilities from the older MPCA system1. The software itself is written in the C language and compiles on a Linux and a Mac OS X environment. The models presented here are a hierarchical extension of discrete component analysis. This is known under many names [2], such as LDA, multi-aspect models, multinomial PCA, etc.