【作者】 李崇威； 杨波； 陈月辉；

【机构】 济南大学信息科学与工程学院；

【摘要】 <正>1 引言人工神经网络(Artificial Neural Networks, ANN)是近年来得到迅速发展的一个前沿课题。由于其大规模并行处理、容错性、自组织和自适应能力和联想功能强等特点,已成为解决很多问题的有力工具,对突破现有科学技术的瓶颈,更深入探索非线性等复杂现象起到了重大作用,已广泛应用到各个工程领域,在各方面取得了很大的进展。更多还原

【Abstract】 Tabu search and neural networks are two significant modern optimization techniques , which are mainly applied to solve difficult optimization problems. As artificial neural networks gain popularity in a variety of application domains,more and more search techniques are applied to optimization problems of neural networks. Tabu search is a quite flexible and promising search technique for obtaining good solutions quickly and has been applied to solve many optimization problems. When the neural networks model is large, training or running it needs taking alot of time on the single CPU and debases the efficiency of operation,maybe thebetter results don’t be obtained. This paper presents a method of using paralleltabu search to optimize neural networks to speed up the implementation of neuralnet-works and to overcome the initial solution dependent of basic tabu search at the same time. The model was evaluated using the data of IDS and obtain the better results.更多还原