logistic-regression

Models the log-odds (see Probabilities) of an event as a linear combination of independent variables. Logarithm function is a perfect loss for fitting a logistic regression curve against binary events - Log loss