Feature selection using tabu search with long-term memories and probabilistic neural networks

摘要: Feature selection is a dimensionality reduction problem in order to reduce measurement costs, shorten computational time, relieve the curse of dimensionality, and improve classification accuracy. In this paper, hybrid approach using tabu search probabilistic neural networks proposed applied feature problems. The algorithm differs from previous research by long-term memory instead short-term avoid necessity delicate tuning length decrease risk generating cycle that traps local optimal solutions. integrated are an outgrowth Bayesian classifiers outperform backpropagation-based their global convergence rapid training. Extensive experiments on real-world data sets performed comparison with indicates can select equal or smaller number features while improving