基于多表达式编程的汇率预测的研究

Exchange Rate Forecasting Using Multi-Expression Programming

由于汇率预测(Exchange Rate Prediction)是一种不确定、非线性、非平稳的时间序列预测问题,传统的方法往往难以得到满意的结果。多表达式编程(Multi Expression Programming,MEP)是一种新型的线性编码的遗传编程(Genetic Programming)的变种。提出一种克隆选择算法优化的多表达式编程模型对国际上3种重要的汇率数据进行了建模和预测,实验结果表明,该模型克服了传统进化算法优化的遗传编程及人工神经网络早收敛、难以找到全局最优解的缺点,取得了令人满意的结果。

Forecasting exchange rate is an uncertain, nonlinear and unstable time - serial prediction problem, which has no satisfying results using traditional method. Multi - Expression programming （MEP） is a new Genetic Programming （GP） variant that uses a linear representation of chromosomes. In this paper a clonal selection -based MEP model for forecasting three major international currency exchange rates is proposed. Empirical results indicate that the proposed method is better than the conventional artificial neural network and genetic programming forecasting models.