三支决策的可视化概念与分析OA

Three-way decision is a novel theory for solving uncertain problems,and enhancing its interpretability is a key research direction.While visualization is a powerful means of achieving interpretability in computer science,a sys-tematic theoretical framework for visualizing three-way decisions has not yet been established,creating an urgent need for highly interpretable visualization tools.Therefore,based on the classic models of three-way decisions,this paper pro-poses a set of methods for visual modeling and reasoning.Firstly,building on a visual representation of the fundamental concepts,this paper addresses the visualization of Decision-Theoretic Rough Sets(DTRS)by constructing a Probability-Cost visualization space.Within this space,the geometric semantics and monotonicity of the three cost-objective func-tions are visually interpreted,which in turn enables a visual derivation of the decision thresholds.Secondly,extending this work to the visualization of Double-Quantitative Three-Way Decisions(DQ-TWD),we further construct a two-di-mensional Relative Conditional Probability-Absolute Conditional Probability space for the visual analysis of the nine double-quantitative decision types.Finally,a case study using a mushroom dataset verifies that the proposed visualiza-tion tools provide effective and interpretable decision support for complex uncertain reasoning applications,demonstrat-ing strong potential for generalization and practical application.