基于som网络-灰关联分析-BP网络预测模型

提出一种基于som网络-灰关联分析-BP网络预测模型，用于股市收盘价的预测．首先采用som神经网络将特性分散的样本划分成不同的子类，然后采用灰关联分析方法，解析了各影响因子对股市收盘价的影响程度，以反映股市收盘价的影响因子作为输入变量，构建了股市收盘价的BP网络预测模型，结果表明：利用灰关联分析，可得出各凶子对股市收盘价影响程度的大小排序，能为股民提供参考；基于BP网络模型的预测结果准确，通过对所采集的股市数据进行测试，证实本文所提出的方法具有有效性．

Combinational Forecasting Model Based on Neural Network and Gray Relational Analysis

A combinational prediction model based on neural network and Gray relational analysis was presenled, which was applied to real-time prediction of economic-technical indexes in stock prices. The proposed method was used to divide samples which have characteristic of decentralization into different sub-classes with the aid of som neural network, and Gray relational analysis method was used to sequence the effects of each factor on price and provide reference for stock-owners. The forecasting results based on neural network were correct, which indicated the proposed method was efficient.