Clustering Techniques: K-means and DBSCAN
摘要
One of the frequent aims of data analysis is grouping the data points into larger clusters or categories of somehow similar objects. For example, we might want to group survey participants according to their sociodemographics, group customers according to their purchasing habits, travelers according to the types of attractions they like to visit, destinations according to their road distance, and so on. Grouping of data records may be the final goal of the analysis, but more frequently, this is just a step leading to an independent analysis of each of the groups or an in-depth study of one of the groups. For example, in the analysis of the University of Florida Gators basketball game attendants survey (Chap. 2 lab), one may pay a particular interest in a deeper analysis of the group likely to bring donations to the team.