Hyperspherical granular computing classification algorithm based on fuzzy lattices

Obtaining changeable granules quickly and accurately is one of the important issues in granular computing. The present work proposes a partial order relation and lattice computing to deal with the aforementioned issue. A hyperspherical granular computing classification algorithm (HSGrCCA) is developed in the framework of fuzzy lattices. HSGrCCA computes a fuzzy inclusion relation between two hyperspherical granules using an inclusion measure function based on a linear positive valuation function induced by the radius of a hyperspherical granule. A fuzzy lattice is formed on the hyperspherical granule set by the dilation operator, erosion operator, and partial order relation. HSGrCCA is trained by introducing control parameter ρρ of the hyperspherical granule size and then obtains changeable hyperspherical granules. Experimental results on machine learning benchmark data sets show that the proposed algorithm increases the generalization ability.