Barakbah, Ali Ridho and Arai, Kohei (2004) Reversed pattern of moving variance for accelerating automatic clustering. EEPIS Journal, 2 (9). pp. 15-21. ISSN 0852-2863
Abstract
This paper proposes a new approach to accelerate a construction of automatic clusters. It initiates an analyse of the moving variance of clusters for each stage of cluster construction, then observes the pattern to find the global optimum as well as avoid the local optima. Using two constraints, valley- tracing and hill-climbing, to find the global optimum, the proposed approach reverses the pattern in order to accelerate the automatic clustering. Experiment result performs the effectiveness of the proposed approach in this paper.
Actions (login required)