An Enhanced Election Algorithm for Data Clustering

Clustering is one of the major data analysis tools needed in many fields such as data mining, machine learning, information retrieval, bioinformatics and pattern recognitionsolutions. Election Algorithm (EA) is an innovative sociopolitically optimization technique which is inspired by the candidates’ behavior in presidential election process. Such as other meta-heuristic algorithms, EA algorithm can be used in different optimization application fields such as machine learning and pattern recognition. This work presents an improved version of EA algorithm (EEA) to data clustering. Compared to canonical EA algorithm, MEA equipped with a migration operator, which improved efficiency of the algorithm. The proposed approach is tested on five well-known datasets from different domains and compared with several baseline methods. The simulation results show that the proposed method outperformed the baseline methods

کلیدواژه ها:

pattern recognition ، data clustering ، meta-heuristic algorithms ، election algorithm ، enhanced election algorithm