基于SAGA—BP神经网络的证券智能分析系统研究

摘要

自然界中生物对复杂事物的处理机制在证券智能分析系统中应用，是提高证券预测分析准确性的一种有效途径。系统的运算功能模块分为两个方面：第一，它结合模拟退火算法和遗传算法的特点，有效的提高了处理复杂问题时算法收敛速度和精度；第二，它使用混合算法对神经网络的进行优化，充分发挥了神经网络的优点，弥补了它的缺点。根据需求对系统的功能模块进行分析设计，文章最后简要介绍MATLAB与Visual C＋＋之间如何通信。%The processing mechanism that the processing of biological in nature for complex things applicationing to securities analysis system is an effective way to improve the accuracy of the prediction of securities analysis. The operation module of the system divided into two aspects： first, the system combined with the characteristics of simulated annealing algorithm and genetic algorithm and effectively improve the algorithm convergence speed and accuracy; second, the system use hybrid algorithm optimized the neural network based on the advantages and disadvantages of the error back propagation neural network. According to user needs, the paper analysised and designed function module of the system and briefly introduced the MATLAB and Visual C＋＋ Communication method.