Machine Learning Guide for Oil and Gas Using Python

Chapter 7: Model evaluation

Abstract

In this chapter, the whole process of model evaluation and improvement is illustrated. The chapter begins with a discussion about the most popular metrics for model assessment. These metrics are depicted for binary classification, multiclass classification, and regression problems with prediction of sand production, rock typing, and PVT estimation examples. After understanding multiple evaluation metrics like precision, accuracy, recall, mean square error, and R², multiple cross-validation techniques for model evaluation are explained. Next, grid search that implements cross-validation to optimize model's hyperparameters and maximize performance metrics is explained with different examples. When a reliable model ...