A Novel Approach for Sequential Three-Way Decision Using Chi-Square Statistic as the Assessment Metric
Remesh Kollezhath Muraleedharan, Latha Ravindran Nair · Revue d intelligence artificielle · 2024
Three-way decisions play a crucial role in addressing decision-making problems in situations of uncertainty.They categorize the decision space into three discrete regions, specifically referred to as the Positive, Negative, and Boundary regions.These models are frequently employed in scenarios that involve the existence of multiple potential alternatives, necessitating the inclusion of a deferral option in addition to the two extremes.This methodology is especially advantageous in situations where the intricacy of the decision-making environment necessitates a more sophisticated examination of potential options that extend beyond a binary choice.The concept of information granularity is the foundation for the ability to conceptualize, comprehend, and apply a sequential three-way decision approach.When working with fine-grained granules, people have the freedom to carefully consider all of their options before deciding on any one.The application of sequential three-way decisions results from the availability of detailed information.Through the application of various techniques and a shift from coarse to fine-grained information granularity, this decision-making method establishes specific thresholds for efficient decision-making.This study introduces an innovative approach for exploring objects that have not made a definitive decision ultimately leading to the convergence of these objects into the Positive and Negative regions.The Chi-square statistic is employed as the evaluative measure for the process of dividing the decision space into three distinct regions.