This chapter describes the main goal of the book, namely the use of evolutionary algorithms to optimize the K-means algorithm. The outline of the book is also given in the chapter.

The chapter describes three phases of a pattern recognition system; data acquisition, feature extraction and classification. In addition, different clustering methods are described.

The chapter describes different components of evolutionary algorithms and what is meant by mathematical optimization. In addition, genetic and evolutionary programming is defined.

The system requirements are divided into two types, functional and nonfunctional. Both of them are described in addition to their data visualization.

In this chapter, the system architecture and different tools that have been used in the implementation process such as Netbeans and JavaFX, are described.

Different methods for visualization of data are described in the chapter. One important aspect is how to handle colours. Colour are separated in three different subchannels. Different methods to reduce the dimension of data are also described.

The chapter gives an introduction to the architecture based on Model-View - Controller (MVC) patterns. Such an architecture separates the user interface from the business logic. The MVC enables us to divide the system into three components. The user provides the parameter values. When an algorithm is terminated the results may be shown graphically.

To compare the different algorithms three data sets have been used. Different benchmarking sets have been applied and the results of the experiments are presented in tables and illustrated graphically.

In this chapter we discuss different challenges of using evolutionary algorithms to optimize the K-means algorithm. One problem is how to handle empty clusters. In addition, the time complexity of the different algorithms is shown.

In this chapter we make a summary of how to optimize the K-means clustering algorithm based on evolutionary computing. The system is still missing a user interface to handle invalid user input. Parallel coordinates that may be used as a tool to visualize data in high-dimensional spaces is only given a short introduction. In addition, Particle Swarm Optimization (PSO) is also mentioned to find global solutions to optimization problems.