神经网络在股票市场预测中的应用

【摘要】 本文分析了人工神经网络的缺陷的特点,提出了基于粒子群优化算法的神经网络模型算法.该算法利用粒子群算法全局寻优能力强等特点,可克服神经网络易陷入局部极小值、训练速度慢的缺陷。通过股票市场的仿真结果表明:基于粒子群算法的神经网络具有较好的全局快速收敛和较强的逼近等性能。更多还原

【Abstract】 The shortcoming of Artificial neural network is analyzed, and the model based on particle swarm optimization algorithm is presented. Using the property that the entire optimizing capability of the algorithm is perfect, the disadvantages that the net training is easily fallen into local minimums and the speed is slow can be avoided. The simulation result shows that rapid global convergence can be obtained by combining particle swarm optimization algorithm and neural network.更多还原