Introduction to Machine learning

Foundation of machine learning

This chapter will focus on the training step. The Upcoming chapter focuses on the validation and testing steps. While there exist many tools and programming libraries that make machine learning methods widely accessible, it’s important that we understand the inner workings of these algorithms and methods so that we can use them correctly and appropriately. Therefore, in this chapter, we’ll lay out important foundation theories before detailing the techniques themselves. In particular, for classification, we’ll discuss:

• K-nearest Neighbor
• Naïve Bayes
• Logistic regression
• Linear discriminant analysis
• support vector machines
• decision trees
• Random forests

For regression, we’ll discuss linear regression and some popular extensions, including ridge and lasso regressions.

To help us learn the practical aspect of using these algorithms, we’ll discuss their technical details. We’ll also augment this discussion with labs in R, where we can practice using the algorithms and setting their parameters with some game data. We’ll use the Dota 2 data in all the labs in this chapter. As with the previous chapters, we’ve included the data with the chapter, which we can access through the lab folder. The chapter includes the following labs:

• Linear Regression Lab
• KNN Lab
• Naive Bayes Lab
• Logistic Regression Lab
• LDA Lab
• SVM Lab
• SVM with Kernel Methods Lab
• DT Lab
• Random Forest Lab