An approach to the construction of event archipelago for chance discovery via genetic algorithm based on correlation matrix

摘要

Chance discovery is a new research topic on cognition psychology inspired deep data analysis. Scene is the contact point of human process and computer process in the double helical model of chance discovery. Event archipelago is the basic foundation for scene construction. Event archipelago and island are defined in this paper. Problem of event archipelago for Chance Discovery based on correlation matrix is a NP hard problem and to solve this problem instantly, GeneticCEA, an approximate algorithm based on genetic algorithm is presented. This paper discusses the performance of GeneticCEA too. Experimental results show that GeneticCEA can run with excellent performance for clustering task while it is treated as a kind of task for the construction of event archipelago in chance discovery.