混沌时间序列预测算法的优度评价

Priority Degree Evaluation of Chaotic Time Series Prediction Algorithm

摘要:

将优度评价方法运用于混沌时间序列预测算法评价中,确定评价算法的指标体系。建立混沌时间序列预测算法关于各评价指标的关联函数来刻画各评价对象的优秀程度并计算关联度。计算出各评价对象的优度值,将其用于对各算法的综合优度的比较。实例证明,该方法直观、实用,可为混沌时间序列预测算法评价提供定量的依据。

Abstract:

The method of priority-degree evaluation was used in the evaluation of chaotic time series prediction algorithm to determine the evaluation index system. The correlation functions on evaluation indexes were built to describe the excellent degrees of evaluation objects in the chaotic time series prediction algorithm, and compute correlation degree. The priority of each object was computed to compare the comprehensive excellence of each algorithm. An example was given to prove that the method was intuitive and practical, and could supply the quantificational basis for the evaluation of chaotic time series prediction algorithms.