produces a predictive score that is a weighted average of the predictors. It finds the optimal set of weights, chosen to maximize the correlation between the composite prediction and the target variable. The optimal weights minimize the MSE (mean squared error) of the predictions
