Kernel-based Fuzzy C-means Clustering Based on Fruit Fly Optimization Algorithm

Fuzzy clustering has emerged as an important tool for discovering the structure of data. Kernel based clustering has emerged as an interesting and quite visible alternative in fuzzy clustering. Aimed at the problems of both a local optimum and depending on initialization strongly in the fuzzy c-means clustering algorithm (FCM), a method of kernel-based fuzzy c-means clustering based on fruit fly algorithms (FOAKFCM) is proposed in this paper. In this algorithm, the fruit fly algorithm is used to optimize the initial clustering center firstly, kernel-based fuzzy c-means clustering algorithm (KFCM) is used to classify data. At the same time we reference classification evaluation index to choose the fuzziness parameter in adaptive way. The clustering performance of FCM algorithm, KFCM algorithm, and the proposed algorithm is testified by test datasets. FCM algorithm and FOAKFCM are used for power load characteristic data classification, respectively. Experiment results show that FOAKFCM algorithm proposed overcomes FCM's defects efficiently and improves the clustering performance greatly.