Multi-feature weighting neighborhood density clustering

Clustering is an important data mining method to discover knowledge and patterns. Feature weighting is widely applied in high-dimensional data mining. In this paper, a multi-feature weighting neighborhood density clustering algorithm is proposed. It uses different dimension reduction algorithms to generate different features, and then, the weights of the features are determined by the discrimination ability. For the clustering algorithm, the center points can be selected by the upper approximation set and lower approximation set. At last, the final clustering result is from the fusion of multiple clustering results. The proposed algorithms and comparison algorithms are executed on the synthetic and real-world data sets. The test results show that the proposed algorithm outperforms the comparison algorithms on the most experimental data sets. The experimental results prove that the proposed algorithm is effective for data clustering.