On graph based clustering algorithms

Clustering problem is one of the important areas of data mining. In the solution of the clustering problem, a data set is intended to have clusters of similar data based on a certain degree of similarity in the same cluster and the clusters as the maximum distance between clusters. In economics, industry, medicine etc. for the clustering problem that has applications in many areas, such as exact-fuzzy, hierarchical-non-hierarchical and so on. There are many different approaches. Another approach to the clustering problem is the graphical basis. In this thesis, different solution approaches for the graphical solution of the clustering problem are investigated. The study of each method on the sample graphical data sets is discussed. In addition, a new graph-based clustering method was examined and detailed analysis was performed on sample data sets.