A modified fuzzy min-max neural network for data clustering and its application on pipeline internal inspection data

In this paper, a modified fuzzy min-max neural network (MFMC) for data clustering is proposed. In MFMC, the centroid information, the similarity and the noise of data are taken into the consideration. What's more, the hyperbox entropy (HE) is first introduced to evaluate the performance of each hyperbox when doing the contraction process. In addition, in order to test the performance of the MFMC model, a series of simulations on benchmark data sets are conducted. Then a real-world application study on the pipeline internal inspection data is also performed. The experimental result indicates that the MFMC has more excellent performance than other existed fuzzy min-max clustering algorithms.