Introduction

Classification is an essential classical problem. The purpose of classification is to automatically classify a text corpus and images into one or more specified classes. In this chapter, we'll use the Keras library to train a classifier. Classification is a general task where machine learning is used, be it question answering, a support request, business question, cancer classification, and so on. This chapter describes how classification problems can be tackled with the open source neural network library. Compared to traditional algorithms, deep learning has several advantages for classification, including flexible models, lower amount of domain expertise being required, and easier ongoing learning.