Bayesian Linear Classifier

Linear binary classification.

This function uses Bayesian inference to find the optimal linear separator in a binary classification problem. A trained model can then be used to make predictions for new samples by computing a probability that that sample belongs to one of the classes.

Error bars are estimated for the regression coefficients.

This classifier will be useful in situations with noisy data or small sample sizes because a parametric linear model will be less prone to overfitting.