摘要
BP神经网络由于具有对数据大规模并行处理及对知识有较强融合能力的优点,应用范围极广.然而也存在一些致命的缺点(如容易陷入局部极小点),通过遗传算法(GA)与BP网络结合,可以有效地解决该问题.优化证券投资组合的仿真模拟实验结果表明,其优化方案比使用二次规划法更优,该方法更具正确性、高效性和实用性.
Due to the BP neural net work＇s capability of massive parallel processing, association memory and approximating non-linear functions, it is applied to wide fields. The problems such as trapping into the local minimum are solved by combining GA and BP neural network. The validity, efficiency and practicability are demonstrated through portfolio investment selection simulation.
出处
《江汉大学学报（自然科学版）》
2005年第3期47-50,共4页
Journal of Jianghan University：Natural Science Edition
