Model performance analysis and model validation in logistic regression

Rosa Arboretti Giancristofaro, Luigi Salmaso

Abstract


In this paper a new model validation procedure for a logistic regression model is presented. At first, we illustrate a brief review of different techniques of model validation. Next, we define a number of properties required for a model to be considered "good", and a number of quantitative performance measures. Lastly, we describe a methodology for the assessment of the performance of a given model by using an example taken from a management study.

Full Text:

pdf (English)


DOI: 10.6092/issn.1973-2201/358