Formation of general type-2 Gaussian membership functions based on the information granule numerical evidence

This paper shows a new technique for forming fuzzy Gaussian membership functions based on the numerical evidence which is found in its information granule. Inspired by the principle of justifiable granularity, and by obtaining a meaningful granule of information, general type-2 Gaussian membership functions are created which better represent a piece of information. Some examples are given, a synthetic example to show the general behavior, as well as an example taken from the iris dataset.