Clustering algorithms
Clustering algorithms are employed to restructure data in somehow ordered subsets so that a meaningful structure can be inferred. A cluster can be defined as a group of data points with some similar features. The way to quantify the similarity of data points is what determines the different categories of clustering.
Clustering algorithms can be divided into different categories based on different metrics or assumptions in which data has been manipulated. We are going to discuss the most relevant categories used nowadays, which are distribution methods, centroid methods, density methods, and hierarchical methods. For each category, a particular algorithm is going to be presented in detail, and we will begin by discussing distribution methods. An example to compare the different algorithms will be discussed, and both the IPython notebook and script are available in the my GitHub book folder at https://github.com/ai2010/machine_learning_for_the_web/tree/master/chapter_2...
