一种求三支决策阈值的模拟退火算法

A Simulated Annealing Algorithm for Learning Thresholds in Three-way Decision-theoretic Rough Set Model

三支决策粗糙集模型作为一种代表性的概率型粗糙集模型,在给定损失函数基础上可以计算出决策所需阈值,能够推导出现有多种概率型粗糙集模型,但是决策粗糙集模型需要合适的先验知识预先设定损失函数,使得三支决策粗糙集在应用过程中受限.基于针对决策粗糙集模型构建的最优化问题,提出了一种模拟退火算法,通过解决该优化问题,能够从数据中学习出三支决策模型所需的阈值.在部分数据集上的实验表明了模拟退火算法在运行时间上要优于现有的算法,基于模拟退火算法求得的阈值能够得到较小的决策风险代价.

Three-way decision-theoretic rough set model is a probabilistic rough set model. It can derive several other probabilistic rough set models by setting corporate cost functions. One limitation of applying the model into more applications is that it needs prior knowledge or expert opinion to get cost functions. This paper gives a simulated annealing algorithm for learning thresholds without any prior knowledge. The algorithm is based on solving an optimization problem with the objective of minimizing decision cost. Compared to the existed algorithm, the experimental result shows its efficiency on running time, and the decision cost made based on the learned thresholds from simulated annealing approach is also less.