灰色预测相对误差在股价区间预测中的应用

Application of Relative Error of Grey Prediction in the Stock Price Interval Prediction

根据相对误差最小的原则对原始数据进行筛选,将G M（1,1）模型用于股票预测,同时利用最小的相对误差和模型预测值将单值预测转换为区间预测,并以中国石化（600027）2012年5~6月3 9个交易数据为例进行了分析。分析结果表明,5 4%的预测区间是有效的。

GM（1,1） model has been taken for stock prediction with the original data selected under the principle of the minimum relative error.Meanwhile,the minimum relative error and model prediction have been used to change single prediction into interval prediction.39 transaction data of Sinopec（600027） from May to June in 2012 have been taken as the example to do the analysis.The results indicate that 54% of the prediction interval is effective.