Abstract
In order to solve the problem that the density peak clustering algorithm (FDP) needs to manually selected the center on the decision graph, an integration algorithm based on FDP and DBSCAN is proposed. Inspired by the FDP algorithm, that is, the data with the largest local density is the first center, and we use DBSCAN algorithm to cluster from the data with the largest local density to get the first category. Then the maximum local density data is found out from the data that has not been clustered, and use the DBSCAN algorithm to find the second category. Repeat this until all data is divided. The proposed algorithm optimizes the DBSCAN algorithm, that is, every iteration starts from the current best point (the point with the largest local density, this is the center of every cluster). Compared with the classical algorithm FDP, FCM, K-means, the proposed algorithm can obtain higher efficiency.