A variety of fuzzy clustering methods exploit the Euclidean distance to quantify resemblance between data points. This distance is effective for revealing spherical clusters, but it does not perform well for data exhibiting more complicated geometry. So, in this paper, we present a new algorithm IFPCM with Minkowski distance applied on real and artificial datasets being generated according to various design factors. We realized a comparison between two algorithms FCM and IFPCM while changing the distance measurement. Finally, we realize a comparison between this new approach on a noisy and incomplete dataset.