Publication

Abstract : Hypergraph learning is the process of representing data in the form of edge-connected graph. This learning process enables data scientists to extract optimal number of consistent features. Hypergraph clustering is an extensive property in the process of feature extraction. Here, clusters are framed with vertices of hypergraph in which vertices in each cluster are strongly connected. In this process, features are represented by vertices of hypergraph and edges are generated with high level of connectivity between features. After forming hypergraph clusters, n number of possible hypergraph subclusters have to be generated based on standard algorithms and finally clusters that contain feasible features are collected. This process can be used in social networking, bioinformatics, and computational intelligent system. In this chapter, we incorporate the principles of rough set dimensionality reduction process with hypergraph. Reduct is the unique term used in Rough set which defines possible subsets from knowledge base. In this chapter, we have proposed the reduct generation process using Rough hypergraph which is implemented in an autistic spectrum disorder dataset.