A classifier is a model that assigns each input to one of a fixed set of categories. In finance the term almost always refers to a machine learning model trained on labeled examples: past card payments marked as genuine or fraudulent, past loans marked as repaid or defaulted, or past trading days marked as up or down. Given a new case, the classifier predicts its label, usually with a probability attached.
Common types include logistic regression, the decision tree, the support vector machine and the artificial neural network. Banks and payment firms use classifiers for fraud detection and credit decisions. Trading firms use them to predict the direction of the next price move. A classification threshold converts the probability into an action. In markets, accuracy alone is a poor measure, because a model that is right slightly more than half the time can still lose money if its errors are larger than its wins.