Derived Multi-population Genetic Algorithm for Adaptive Fuzzy C-Means Clustering

Fuzzy C-Means (FCM) is a common data analysis method, but the clustering effect of this algorithm easily affected by initial centers. Currently, scholars often use multiple population genetic (MPGA) to optimize centers, MPGA has insufficient global search ability and lacks self-adaptability, prone premature convergence, poor Therefore, paper proposes an adaptive FCM DMGA-FCM based on derivative (DMGA). In algorithm, firstly, operator, which proposed for first time in paper, performs operations initialized populations improve algorithm's searchability deal with lack inter-population ability. Secondly, probability fuzzy control operator used dynamically adjust adaptability turn enhances merit-seeking DMGA avoids convergence. Finally, center optimized enhance algorithm. The simulation experiments MRI brain map application results show that can obtain better medical image segmentation compared other related algorithms.