Quantum Clustering Ensemble

• Main Authors: Peizhou Tian, Shuang Jia, Ping Deng, Hongjun Wang
• Format: Article
• Language: English
• Published: Atlantis Press 2020-11-01
• Series: International Journal of Computational Intelligence Systems
• Subjects: Clustering ensemble; Quantum clustering ensemble; Base clustering
• Online Access: https://www.atlantis-press.com/article/125946534/view

Clustering ensemble combines several base clustering results into a definitive clustering solution which has better robustness, accuracy, and stability, and it can also be used in knowledge reuse, distributed computing, and privacy preservation. In this paper, we propose a novel quantum clustering ensemble (QCE) technique derived from quantum mechanics. The idea is that basic labels are associated with a vector in Hilbert space, and a scale-space probability function can be constructed for clustering ensemble. In detail, an operator in Hilbert space is represented by the Schrodinger equation of the probability function as a solution. Firstly, the base clustering results are regarded as new features of the original dataset, and they can be transformed into Hilbert space as vectors. Secondly, a QCE model is designed and the corresponding objective function is illustrated in detail. Furthermore, the objective function is inferred and optimized to obtain the minimum result, which is then used to determine the centers. At last, 5 base clustering algorithms and 5 clustering ensemble algorithms are tested on 12 several datasets for comparing experiments, and the experimental results show that the QCE is very competitive and outperforms the state of the art algorithms.