Abstract:

Density-based clustering algorithms have recently gained in popularity in data mining because of their ability to discover clusters of any shape. For a density based clustering algorithm, clusters are considered to be high density regions separated from each other by hollow regions or regions of low density. Objects located in low density regions are generally considered noise or outliers. There are several methods of classification based on density; the choice of a particular method will depend on the type of data and the size of the data set. The treated data is with increasing volume and complexity, which gives birth to new challenges for classification algorithms, in particular density-based clustering algorithms. In this paper, we focus on density-based classification approach. We present the principle, the basic concepts as well as the challenges for the classification based on density and we cite by the way the reference algorithms in this context. We end this paper with a summary of all of the cited algorithms.