Programming Collective Intelligence

Chapter 12. Algorithm Summary

This book has introduced a number of different algorithms, and if you’ve been working through the examples, you now have Python code that implements many of them. The earlier chapters are structured around working through an example problem with algorithms and variations introduced throughout the chapter. This chapter will be a reference for the algorithms covered, so when you want to do some data mining or machine learning on a new dataset, you can look at the algorithms here, decide which one is appropriate, and use the code you’ve already written to analyze your data.

To save you from going back through the book to find the details of an algorithm, I’ll provide a description of each one, a high-level overview of how it works, what sort of datasets you can apply it to, and how you would use the code you’ve previously written to run it. I’ll also mention some of the strengths and weaknesses of each algorithm (or, if you like, how to sell the idea to your boss). In some cases, I’ll use examples to help explain the properties of the algorithm. These examples are greatly simplified—most are so simple you can solve them just by looking at the data yourself—but they are useful for illustration.

Supervised learning methods, which guess a classification or a value based on training examples, will be covered first.

Bayesian Classifier

Bayesian classifiers were covered in Chapter 6. In that chapter, you saw how to create a document classification system, such as ...