Bayesian information criterion

Bayesian Information Criterion, or BIC, is a way of helping computers figure out which of several possible models best describes a set of data. Think of a model like a big equation with lots of numbers that represent different things. What BIC does is compare two or more possible equations to see which one explains the data best. It does this by looking at how much information each equation uses. The equation that uses the least amount of information is usually the best one.