Model Simplification for Supervised Classification of Metabolic Networks

Many real applications require the representation of complex entitiesand their relations. Frequently, networks are the chosen data structures, due totheir ability to highlight topological and qualitative characteristics. In this work,we are interested in supervised classication models for data in the form of net-works. Given two or more classes whose members are networks, we build math-ematical models to classify them, based on various graph distances. Due to thecomplexity of the models, made of tens of thousands of nodes and edges, we focuson model simplication solutions to reduce execution times, still maintaining highaccuracy. Experimental results on three datasets of biological interest show theachieved performance improvements.