Intuitionistic fuzzy multi-attribute decision-making method and application of matlab based on foreground theory and gray target decision

1. Prospect theory
The prospect theory proposed by Kahneman and Tversky amends the traditional utility theory of expected utility and constructs a new decision framework model. It assumes that the insurance decision-making process consists of two stages: editing and evaluation. In the first stage, individuals collect and process data by means of frames, reference points, etc. In the second stage, according to the value function and weighting function To make decisions and judgments, and at the same time analyze the reasons for the "S-shaped" change in the value function under the two scenarios of gains and losses.
In prospect theory, the prospect value is determined by both the value function and the weight function. The reference point for decision-making should be selected first (the following formula assumes that the reference point is the origin of coordinates), then:

2. The decision step
first obtains the uncertainty Alternative information and attributes and other relevant information under the conditions, and then normalize the attribute values ​​of different data types, and convert them into several intuitionistic fuzzy decision matrices under different state conditions, using the intuitionistic fuzzy score function to simplify the decision matrix Then, the gray target decision method is used to establish positive and negative bullseyes, and the three reference point matrices of the model are established in conjunction with the average expected value. The information of the intuitionistic fuzzy decision matrix is ​​based on the three reference points for information aggregation based on prospect theory, and then the weighted average The child will integrate the three prospect matrices into a comprehensive prospect value decision matrix for each alternative under different attribute conditions. Then, use the optimization model that maximizes the prospect value to calculate the attribute weights. Sort.