Simple machine learning algorithms implemented in Python.
As of 2012-11-24, it has a naive (gaussian) bayes, a binary logistic regression, and an adaBoosted stump decision (the stump decision can be used independently, if that's what you're into).
Here are some of their characterisitcs and limitations:
- Naive Bayes: multinomial classification, the targets can be any integer.
- Logit Reg.: binary classification, the targets must be 0 and 1.
- Boosted stump: binary classification, the targets must be -1 and 1.