Implementations with DL4J

ND4J is the library that helps you to implement deep learning easily and conveniently. However, you have to implement the algorithms yourself, which is not too different from its implementation in the previous chapters. In other words, ND4J is just a library that makes calculating numerical values easier and is not a library that is optimized for deep learning algorithms. One library that makes deep learning easier to handle is DL4J. Fortunately, as for DL4J, some example code with typical methods is published on GitHub (https://github.com/deeplearning4j/dl4j-0.4-examples). These examples are used on the premise that you are using DL4J's version 0.4-*. When you actually clone this repository, please check the latest version again. In this section, we'll extract the fundamental part from these sample programs and take a look at it. We'll reference the forked repository on https://github.com/yusugomori/dl4j-0.4-examples as a screenshot in this section.

Setup

Let's first...