The analysis and research of clustering algorithm based on PCA
Ni Liu, Si Jinhang · 2017
Clustering analysis has been applied in many fields as a key technology in data analysis and processing. A lot of clustering algorithms which have their own advantages and disadvantages have been presented by many scholars. A clustering algorithm framework based on the principal component analysis is presented in this paper. Three steps are taken in the clustering algorithm such as following. Firstly, the data set are dealt with by the way of the principal component analysis (PCA), then the main components are selected to construct the new data set according to the analysis results, finally, the new data set is clustered by the way of DBSCAN. In the algorithm, the dimensions of the new data set are reduced. On account of the lower dimension data set, the amount of calculation is greatly reduced so as to improve the efficiency of the algorithm. The analyzing results of the algorithm simulation and conclusion are also given in this paper. The numerical simulations show that our algorithm is suitable for the occasion on which the clusters have the obvious difference between each other and this clustering algorithm can produce perfect clustering effect for the data set which is extracted by the principal component.