Logistic regression is one of the most popular methods for predicting binary or multi-class outcomes. In situations involving binary targets it yields a linear prediction function that is transformed to produce predicted probabilities of response for scoring observations and coefficients that are easily transformed into odds ratios, which are useful measures of predictor effects on response probabilities. For unordered (nominal) multi-class outcomes, it produces a set of linear prediction functions, each of which is used to calculate odds of responding in a given category relative to a chosen reference category.

Next steps

Like your visualization? Why not deploy it? For more information, see Deploy a model.