CHAPTER 7Building Ensemble Models with Python

This chapter uses several available Python packages to build predictive models using the ensemble algorithms that you saw in Chapter 6, “Ensemble Methods.” The problems used to illustrate them were introduced in Chapter 2, “Understand the Problem by Understanding the Data.” You saw in Chapter 5, “Building Predictive Models Using Penalized Linear Methods,” how to build predictive models for them using penalized linear regression. This chapter uses ensemble methods to solve the same problems. That will enable you to compare the algorithms and the available Python packages in terms of how easy the packages are to use, what kinds of accuracy is achievable with ensemble methods versus penalized linear regression, how the training times compare, and so on. The end of the chapter shows some summary comparisons of the various algorithms you’ve become familiar with.

Solving Regression Problems with Python Ensemble Packages

The next several sections demonstrate the application of available Python packages for building ensemble models. You will see the things you learned in Chapter 6 in action. The methods explained in Chapter 6 will be used on the series of problems explored in Chapter 2 and then used to demonstrate the application of penalized linear regression in Chapter 5. Using the same problems makes it possible to compare the algorithms covered here along several dimensions, including raw performance, training time, and ease of use. ...

Get Machine Learning in Python: Essential Techniques for Predictive Analysis now with O’Reilly online learning.

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