1. Random Walk Quantum Clustering Algorithm Based on Space
- Author
-
Xiao Shufen, Hong-Yang Ma, and Yu-Min Dong
- Subjects
Physics and Astronomy (miscellaneous) ,Computer science ,General Mathematics ,k-means clustering ,Quantum simulator ,Quantum entanglement ,Function (mathematics) ,Random walk ,01 natural sciences ,010305 fluids & plasmas ,Data point ,Mathematics::Probability ,0103 physical sciences ,Quantum walk ,010306 general physics ,Cluster analysis ,Algorithm - Abstract
In the random quantum walk, which is a quantum simulation of the classical walk, data points interacted when selecting the appropriate walk strategy by taking advantage of quantum-entanglement features; thus, the results obtained when the quantum walk is used are different from those when the classical walk is adopted. A new quantum walk clustering algorithm based on space is proposed by applying the quantum walk to clustering analysis. In this algorithm, data points are viewed as walking participants, and similar data points are clustered using the walk function in the pay-off matrix according to a certain rule. The walk process is simplified by implementing a space-combining rule. The proposed algorithm is validated by a simulation test and is proved superior to existing clustering algorithms, namely, Kmeans, PCA + Kmeans, and LDA-Km. The effects of some of the parameters in the proposed algorithm on its performance are also analyzed and discussed. Specific suggestions are provided.
- Published
- 2018