Effective Amazon Machine Learning

Exploring the model performances

You may recall from Chapter 2: Machine Learning Definitions and Concepts, that in a binary classification context, a logistic regression model calculates for each sample to be predicted a probability — the probability of belonging to one class or the other. The model will not directly output the class of the sample to be predicted. The sample is assigned to one class or the other depending on whether the probability is below or above a certain threshold. By default, this threshold is set to 0.5. Although the AUC score given by the evaluation does not depend on the value of the decision threshold, other classification metrics do. We can change the value of the threshold and see how that impacts our predictions. ...