some kind of blog

I needed a Bayesian classifier with support for some custom add-ons. The implementation that is in the script provided here has been done in Python. It reads data from tab-separated textfiles or alternatively supports reading from MySQL databases. Additionally, it also has the possibility to include combinations of features as orthogonal sparse bigrams (OSBs). The classifier gives the scores and class labels for a user-definable number of classes. The source code is well documented with comments. There is no usage() function as of yet, so the command line arguments are only documented within the source.
(more…)