Ecoli Veri Protein Lokalizasyonunda Bulanık ve Olabilirlikli Kümeleme Algoritmalarının Analizi

Abstract

Clustering is a process of dividing the objects into subgroups so that the same set of data is similar, but the data of different clusters is different. The basis of the fuzzy clustering algorithms is the C-Means families and the strongest algorithm is the Fuzzy C-means (FCM) algorithm. In this study; FCM, Possibilistic Fuzzy C-means (PFCM), Fuzzy Possibilistic C-means (FPCM) and Possibilistic C-means (PCM) algorithms are used to classify the several real data sets which are E.coli, wine and seed data sets into different clusters by MATLAB program. Also, the results of PFCM, FPCM, PCM and FCM algorithms are compared according to the classification accuracy, root mean squared error (RMSE) and mean absolute error (MAE). The results show that the PFCM and FPCM algorithms have better performance than FCM and PCM according to criteria for comparing the performances.