TY - JOUR AU - Szilágyi, László AU - Lefkovits, Szidónia AU - Szilágyi, Sándor Miklós TI - Self-Tuning Possibilistic c-Means Clustering Models JF - INTERNATIONAL JOURNAL OF UNCERTAINTY FUZZINESS AND KNOWLEDGE-BASED SYSTEMS J2 - INT J UNCERTAIN FUZZ VL - 27 PY - 2019 SP - 142 EP - 158 PG - 17 SN - 0218-4885 DO - 10.1142/S0218488519400075 UR - https://m2.mtmt.hu/api/publication/31056733 ID - 31056733 N1 - Export Date: 4 August 2021 CODEN: IJUSF Funding details: Magyar Tudományos Akadémia, MTA Funding text 1: This research was supported by the János Bolyai Research Fellowship program of the Hungarian Academy of Sciences. AB - The relaxation of the probabilistic constraint of the fuzzy c-means clustering model was proposed to provide robust algorithms that are insensitive to strong noise and outlier data. These goals were achieved by the possibilistic c-means (PCM) algorithm, but these advantages came together with a sensitivity to cluster prototype initialization. According to the original recommendations, the probabilistic fuzzy c-means (FCM) algorithm should be applied to establish the cluster initialization and possibilistic penalty terms for PCM. However, when FCM fails to provide valid cluster prototypes due to the presence of noise, PCM has no chance to recover and produce a fine partition. This paper proposes a two-stage c-means clustering algorithm to tackle with most problems enumerated above. In the first stage called initialization, FCM with two modifications is performed: (1) extra cluster added for noisy data; (2) extra variable and constraint added to handle clusters of various diameters. In the second stage, a modified PCM algorithm is carried out, which also contains the cluster width tuning mechanism based on which it adaptively updates the possibilistic penalty terms. The proposed algorithm has less parameters than PCM when the number of clusters is c > 2. Numerical evaluation involving synthetic and standard test data sets proved the advantages of the proposed clustering model. LA - English DB - MTMT ER - TY - JOUR AU - Gosztolya, Gábor AU - Szilágyi, László TI - Application of fuzzy and possibilistic c-means clustering models in blind speaker clustering JF - ACTA POLYTECHNICA HUNGARICA J2 - ACTA POLYTECH HUNG VL - 12 PY - 2015 IS - 7 SP - 41 EP - 56 PG - 16 SN - 1785-8860 DO - 10.12700/aph.12.7.2015.7.3 UR - https://m2.mtmt.hu/api/publication/2992199 ID - 2992199 N1 - MTA-SZTE Research Group on Artificial Intelligence of the Hungarian Academy of Sciences and University of Szeged, Tisza Lajos krt. 103, Szeged, H-6720, Hungary Dept. of Control Engineering and Information Technology, Budapest University of Technology and Economics, Hungary Computational Intelligence Research Group, Dept. of Electrical Engineering, Sapientia University of Transylvania, Tîrgu Mureş, Romania Cited By :8 Export Date: 12 February 2021 Funding details: PD103921 LA - English DB - MTMT ER - TY - JOUR AU - Szilágyi, László AU - Szilágyi, Sándor Miklós TI - Generalization rules for the suppressed fuzzy c-means clustering algorithm JF - NEUROCOMPUTING J2 - NEUROCOMPUTING VL - 139 PY - 2014 SP - 298 EP - 309 PG - 12 SN - 0925-2312 DO - 10.1016/j.neucom.2014.02.027 UR - https://m2.mtmt.hu/api/publication/2698866 ID - 2698866 N1 - Sapientia University of Transylvania, Faculty of Technical and Human Sciences, 540485 Tîrgu Mureş, Şoseaua Sighişoarei 1/C, Romania Budapest University of Technology and Economics, Department of Control Engineering and Information Technology, H-1117 Budapest, Magyar tudósok krt. 2, Hungary Petru Maior University, Department of Informatics, 540088 Tîrgu Mureş, Str. Nicolae Iorga Nr. 1, Romania Export Date: 13 March 2025; Cited By: 31; Correspondence Address: S.M. Szilágyi; Petru Maior University, Department of Informatics, 540088 Tîrgu Mureş, Str. Nicolae Iorga Nr. 1, Romania; email: szsandor72@yahoo.com; CODEN: NRCGE AB - Intending to achieve an algorithm characterized by the quick convergence of hard c-means (HCM) and finer partitions of fuzzy c-means (FCM), suppressed fuzzy c-means (s-FCM) clustering was designed to augment the gap between high and low values of the fuzzy membership functions. Suppression is produced via modifying the FCM iteration by creating a competition among clusters: for each input vector, lower degrees of membership are proportionally reduced, being multiplied by a previously set constant suppression rate, while the largest fuzzy membership grows to maintain the probabilistic constraint. Even though so far it was not treated as an optimal algorithm, it was employed in a series of applications, and reported to be accurate and efficient in various clustering problems. In this paper we introduce some generalized formulations of the suppression rule, leading to an infinite number of new clustering algorithms. Further on, we identify the close relation between s-FCM clustering models and the so-called FCM algorithm with generalized improved partition (GIFP-FCM). Finally we reveal the constraints under which the generalized s-FCM clustering models minimize the objective function of GIFP-FCM, allowing us to call our suppressed clustering models optimal. Based on a large amount of numerical tests performed in multidimensional environment, several generalized forms of suppression proved to give more accurate partitions than earlier solutions, needing significantly less iterations than the conventional FCM. LA - English DB - MTMT ER - TY - CHAP AU - Szilágyi, László ED - Torra, V ED - Narakawa, Y ED - Yin, J ED - Long, J TI - Fuzzy-possibilistic product partition: A novel robust approach to c-means clustering T2 - Modeling Decisions for Artificial Intelligence PB - Springer Netherlands CY - Heidelberg CY - Berlin SN - 9783642225888 T3 - Lecture Notes in Computer Science, ISSN 0302-9743 ; 6820. PY - 2011 SP - 150 EP - 161 PG - 12 DO - 10.1007/978-3-642-22589-5_15 UR - https://m2.mtmt.hu/api/publication/3344749 ID - 3344749 N1 - Export Date: 13 March 2025; Cited By: 26; Correspondence Address: L. Szilágyi; Faculty of Technical and Human Science, Sapientia - Hungarian Science University of Transylvania, Tîrgu-Mureş, Romania; email: lalo@ms.sapientia.ro; Conference name: 8th International Conference on Modeling Decisions for Artificial Intelligence, MDAI 2011; Conference date: 28 July 2011 through 30 July 2011; Conference code: 85898 AB - One of the main challenges in the field of c-means clustering models is creating an algorithm that is both accurate and robust. In the absence of outlier data, the conventional probabilistic fuzzy c-means (FCM) algorithm, or the latest possibilistic-fuzzy mixture model (PFCM), provide highly accurate partitions. However, during the 30-year history of FCM, the researcher community of the field failed to produce an algorithm that is accurate and insensitive to outliers at the same time. This paper introduces a novel mixture clustering model built upon probabilistic and possibilistic fuzzy partitions, where the two components are connected to each other in a qualitatively different way than they were in earlier mixtures. The fuzzy-possibilistic product partition c- means (FP3CM) clustering algorithm seems to fulfil the initial requirements, namely it successfully suppresses the effect of outliers situated at any finite distance and provides partitions of high quality. LA - English DB - MTMT ER -