【摘要】 提出了利用贝叶斯正则化BP神经网络对股票指数进行预测。通过对比实验表明,贝叶斯正则化的BP神经网络比相同条件下采用其他改进算法有较好的泛化能力,对股票指数预测有很好的效果。更多还原

【Abstract】 Using Bayesian regularization BP neural network to prodict stock index was proposed.Experimental results show that Bayesian regularization method has stronger generalization capabilities than other improved BP algorithms under the same situation,and is well effective to forecast the stock index.更多还原