基于GEP方法的数据流预测模型

传统的预测建模方法有曲线拟合、线性回归分析等，这些方法通常只适用于求解结构简单的多项式函数。该文采用基因表达式程序设计方法，该算法简便、易于遗传操作，并且其搜索空间广阔，函数复杂度高，能广泛适用于各种类型的数据流预测。在此基础上，提出当预测模型失效时的大变异策略，收到了很好的效果。

Data Stream Prediction Model Based on GEP Method

Many traditional methods in the field of forecasting, including curve simulation, linear regression, etc, which are applied only to solve simple polynomial functions. Adopting gene expression programming (GEP), this paper proposes a predictive mathematical model for forecasting the aggregatde value over data streams. The algorithm is simple and easy to operate which search functions in the great space. As a result, this forecasting model can be used in many kinds of the data stream. When the frequency of forecast failing is greater than a predefined threshold, an adaptive strategy for the predictive mathematical model is proposed.