I have some issues understanding how a likelihood of an event is predicted with an OLS regression. For example, if the outcome in the training set is a binary variable and a sta
