Optimized feature selection using NeuroEvolution of Augmenting Topologies (NEAT)

摘要: In most real-world problems, we are dealing with large size datasets. Reducing the number of irrelevant/redundant features dramatically reduces running time a learning algorithm and leads to more general concept. this paper, realization feature selection through NeuroEvolution Augmenting Topologies (NEAT) [1] is investigated which aims pick subset that relevant target Two major goals in machine discovery improvement solutions complex problems. Complexification, incremental elaboration adding new structure, achieves both these goals. Hence, work, power complexification NEAT method demonstrated evolves increasingly neural network architectures. When compared evolution networks fixed discovers significantly sophisticated strategies. The results show can provide better accuracy result than conventional MLP improve accuracy.