Metaheuristic Algorithms in Fuzzy Clustering
Sourav De, Sandip Dey, Siddhartha Bhattacharyya · 2020
Fuzzy clustering suffers from the fact that the number of clusters in the given dataset should be known beforehand. It is also sensitive to noise and outliers. The local search-based algorithms have been widely used in solving several clustering problems. This chapter presents a simulated annealing-based clustering problem, which can be efficiently handled and solved using the metaheuristic clustering algorithms. The ant colony optimization (ACO) algorithm is applied to overcome the shortcomings of the fuzzy clustering methods. Image segmentation is done using ACO-based fuzzy clustering. To overcome the shortcomings of Fuzzy c-means (FCM), a new fuzzy subspace clustering algorithm based on improved firefly algorithms is presented. In this approach, the global optimization capability of the firefly algorithm, strong local search features of FCM, and learning calculation for feature weights of reliability-based fuzzy clustering-means are taken into consideration.