A Novel K-Means Clustering Algorithm with a Noise Algorithm for Capturing Urban Hotspots

With the development of cities, urban congestion is nearly an unavoidable problem for almost every large-scale city. Road planning effective means to alleviate congestion, which a classical non-deterministic polynomial time (NP) hard problem, and has become important research hotspot in recent years. A K-means clustering algorithm iterative analysis that been regarded as solve road problems by scholars past several decades; however, it very difficult determine number clusters sensitively initialize center cluster. In order these problems, novel based on noise developed capture hotspots this paper. The employed randomly enhance attribution data points output results adding judgment automatically obtain given Four unsupervised evaluation indexes, namely, DB, PBM, SC, SSE, are directly used evaluate analyze results, nonparametric Wilcoxon statistical method verify distribution states differences between results. Finally, five taxi GPS datasets from Aracaju (Brazil), San Francisco (USA), Rome (Italy), Chongqing (China), Beijing (China) selected test effectiveness proposed comparing with fuzzy C-means, K-means, plus approaches. compared experiment show can reasonably cluster, demonstrates better performance accurately obtains well effectively capturing hotspots.