4. Proposed algorithms
The IDPSO algorithm (Zhang et al., 2013) introduces ωl,c1lωl,c1l and c2lc2l adaptive weights for each particle l, with the goal of improving the performance of PSO. Since it AZD5438 achieved promising results in function optimization problems ( Zhang et al., 2013), a hybrid between FCM and IDPSO (FCM–IDPSO), may be a good alternative to FCM–PSO to solve the fuzzy clustering problem. Eq.