Chapter 7 Evaluating Model Performance

In this chapter we shall explore the process of evaluating the results delivered by a machine learning algorithm and offer various techniques for obtaining better model performance. When we speak of assessing how well a model performs, we generally think of fit measure (R2, adjusted R2, RMSE, etc.), but what we really would like to know is how well a particular model predicts based on new data. This notion finds its basis in the scientific method where observation yields description and experimentation yields explanation—using statistical learning models with the goal of explanation and/or ...

Get Machine Learning and Data Science: An Introduction to Statistical Learning Methods with R now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.