Quantum K-Medians Algorithm Based on Hamming Distance
With the advent of the era of big data,traditional data similarity measurement algorithms are no longer suitable for clustering of high-dimensional data.Based on quantum computing,control-swap gate is proposed to cal-culate the similarity between data.However,it is difficult to decompose and prepare the initial quantum state,which reduces the practicability of the algorithm.Therefore,in this paper proposed a quantum k-medians algorithm based on Hamming distance(QHk-medians)to cluster high-dimensional structured data,which is mainly composed of subrou-tines Hamm DistCalc and GroverOptim.We designed the universal quantum circuit of the subroutines Hamm DistCalc and GroverOptim,and conducted an experimental simulation based on IBM Qiskit.The proposed QHk-medi-ans algorithm has a simple quantum state construction and can accurately measure the similarity between two inde-pendent high-dimensional data and achieve clustering.Compared with classical k-media,the algorithm has lower time complexity and higher classification accuracy.
NETL
NSTL
万方数据
维普
