摘要
介绍了人工神经网络的基本原理和传统股市预测方法的缺点,建立BP神经网络模型,以某个股实际收盘价为原始数据样本,对网络进行训练后,对股票价格进行了短期预测,并计算出预测值和实际值的误差。通过实验发现该模型收敛速度快,预测精度高。
This article introduced the nerve network basic concept and the shortcoming of the traditional method of the stock market forecasts, establishes the BP nerve network model, takes some stock actual closing price as the primary data sample, carries on the training after the network, has carried on the short-term forecast to the stock price, and calculates the forecast value and the actual value error. Through the experiment discovered this model convergence rate quick, the forecast precision is high.
出处
《微电子学与计算机》
CSCD
北大核心
2007年第11期147-151,共5页
Microelectronics & Computer
