Research on Data Fusion Method of Mathematical Creativity Education based on AHP Hierarchical Analysis

Abstract

Dempster-Shafer (DS) evidence theory is widely employed in the real world as a primary instrument for modeling uncertainty. However, when applying Dempster's combination rule, seemingly conflicting pieces of data could be merged to yield surprising results. There are a number of proposed scales for measuring the level of discrepancy between pieces of evidence. However, the presented methods only use a single metric to analyze the conflicting results. However, it is usually unwise to rely on a single criterion to measure the extent to which the facts conflict with one another. For the reason that flaws, differences, differences, and ambiguity all contribute to a higher level of conflict between the evidence. The objective of this study is to propose a novel data fusion method based on AHP for enhancing mathematical creativity education. This work stands out by effectively integrating multiple criteria to improve decision-making processes in educational settings. Given the efficacy of the approach proposed in this study in bringing together seemingly disparate data, we want to eventually extend its application to other forms of uncertainty theory, such as fuzzy set theory and imprecise probabilities. The concept of direct inquiry into the total ambiguity of evidence within the context of discernment will also be explored. Several criteria factors are used to determine the level of disagreement between the information presented here. A proposed analytic hierarchy technique uses multiple criteria to properly weight each piece of data. The initial stage is to assign a numerical value to each piece of evidence based on how well it satisfies each of the criteria. The criterion layer's covariance matrix can be derived by examining the relationship between the criteria's numerical values. After collecting data on each criterion's quantitative change, a fuzzy preference relation matrix can be built. The scheme layer makes use of a fuzzy preference relation matrix in place of a traditional pairwise comparison matrix. Each piece of evidence is given an overall weight based on its combined scheme weight and criteria weight. Two numerical experiments are offered to illustrate the effectiveness of the proposed method after the final weights have been applied to the primary evidence and the evidence has been combined using Dempster's rule. Results show that the proposed methodology outperforms alternative approaches to dealing with contradictory evidence discussed in the literature.