Research On TOPSIS Multi-attribute Decision Making Based On Distance Computation Improvement

This is the era of information explosion,information promotes the rapid development of global economic integration.With the increasingly fierce market competition,in order to cope with the pressure of this fierce competition,every step of the decision made by enterprises is very important.For example,the comprehensive performance evaluation of a certain product,the profitability of the enterprise,the debt ability of the enterprise and other decision-making issues.A concise multi-attribute decision making method is needed to evaluate the research.TOPSIS method is a simple multi-attribute decision making method,which is generally used to solve the optimal ranking problem of various objects,such as the evaluation of the product quality of the enterprise and the comprehensive evaluation and optimization of the financial affairs of the enterprise.This has played a vital role in the management of managers and the progress of enterprises.However,the traditional TOPSIS method ignores the correlation between the evaluation indexes.when there is the phenomenon of attribute correlation,the Euclidean distance will fail and the results will deviate,which will lead to the inaccurate evaluation results.The focus of this paper is to construct an intelligent decision model with strong decision ability by using Mahalanobis distance,cosine similarity distance and vertical distance.Mahalanobis distance instead of Euclidean distance solves the problem related to traditional TOPSIS attributes,but the application of Mahalanobis distance requires a prerequisite:the number of alternatives is greater than or equal to the number of decision indicators.When this premise can not be satisfied,cosine similarity instead of Euclidean distance has achieved good results.In this study,on the basis of discussing the low effectiveness of the traditional vertical distance,the multi-dimensional complex problem is simplified to a three-dimensional problem by using the connection vector.Vertical distance can intuitively measure the distance between the evaluation object and the ideal solution in three-dimensional space,thus simplifying the complex problem into a problem of solving geometric distance and improving the effectiveness of vertical distance.Finally,the mechanism of the improved method is explained through a specific case analysis.