广义多尺度决策表最优尺度选择的快速算法  

A fast algorithm for optimal scale selection of generalized multi-scale decision tables

作　　者：张晓燕[1] 黄雨阳 ZHANG Xiaoyan;HUANG Yuyang(School of Artificial Intelligence,Southwest University,Chongqing 400715,China)

机构地区：[1]西南大学人工智能学院,重庆400715

出　　处：《闽南师范大学学报（自然科学版）》2024年第2期1-7,共7页Journal of Minnan Normal University：Natural Science

基　　金：国家自然科学基金项目(12371465);重庆市自然科学基金项目(CSTB2023NSCQ-MSX1063)。

摘　　要：广义多尺度决策表的条件属性和决策属性都具有多个尺度.最优尺度能够将较粗的条件属性与较细的决策属性相结合,达到效率与精度的平衡.然而现有的最优尺度选择算法计算效率较低.为此,提出一种最优尺度选择的快速算法.首先探讨最优尺度的一些性质和边界域的变化情况,给出判断边界域是否相等的条件以及最优尺度的等价定义;然后一种快速算法.最后通过数值实验表明其相较于现有算法速度更快,能有效解决最优尺度选择问题,该算法在计算效率方面取得了显著的提升,实现了在较短的时间内得出最佳结果的目标.Each attribute of generalized multi-scale decision tables has multiple scales,whether it is a conditional attribute or a decision attribute.The optimal scale combines the coarser condition attributes with the finer decision attributes,so as to achieve a balance between efficiency and accuracy.However,the existing optimal scale selection algorithms lack computational efficiency.In this paper,a fast algorithm for optimal scale selection is proposed.First,some properties of the optimal scale and the variation of boundary region are discussed.Second,the conditions for judging the equality of boundary region and the equivalent definition of the optimal scale are given.Then a fast algorithm is proposed.Finally,numerical experiments show that the algorithm is faster than the existing algorithm and can effectively solve the optimal scale selection problem.The algorithm has achieved a significant improvement in computational efficiency.And it achieves the goal of obtaining the best results in a relatively short time.

关 键 词：最优尺度选择 粒计算 粗糙集 广义多尺度决策表 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]