Implementing machine learning algorithms by yourself is probably the best way to learn machine learning, but you can progress much faster if you step on the shoulders of the giants and leverage one of the existing open source libraries.

This chapter reviews various libraries and platforms for machine learning in Java. The goal is to understand what each library brings to the table and what kind of problems it is able to solve.

In this chapter, we will cover the following topics:

• The requirement of Java for implementing a machine learning application
• Weka, a general purpose machine learning platform
• The Java machine learning library, a collection of machine learning algorithms
• Apache Mahout, a scalable machine learning platform
• Apache Spark, a distributed machine learning library
• Deeplearning4j, a deep learning library
• MALLET, a text mining library

We'll also discuss how to design the complete machine learning application stack for both single-machine and big data apps by using these libraries with other components.