ch_10

Chapter 10: Making Better Predictions – Optimizing Models

This chapter shows you strategies for improving the performance of your machine learning models.

Content

We will be working in the following notebooks, which each build upon their corresponding notebooks in chapter 9:

• planets_ml.ipynb: shows how to use grid search to tune model hyperparameters, gather feature importances, and use regularization with the linear regression model using planet data from the previous chapter
• red_wine.ipynb: walks through hyperparameter tuning with grid search, generating polynomial and interaction features, creating feature unions, and building ensemble models with the red wine quality classifier from the previous chapter as a starting point
• wine.ipynb: discusses dimensionality reduction, decision trees, and error analysis using the white or red wine classifier built in the previous chapter

All the datasets necessary for the aforementioned notebooks, along with information on them, can be found in the data/ directory. The end-of-chapter exercises will also use this data to practice building and tuning ML models with scikit-learn; solutions to these exercises can be found in the repository's solutions/ch_10/ directory.