1. Analytic Hierarchy Process
The Analytical Hierarchy Process (AHP) divides indicators into hierarchies, decomposes complex problems into a series of indicators based on the nature of the problem and the overall goal to be achieved, and divides them into 逻辑different levels according to relationships, thus forming a hierarchical structure.
Then through 两两比较the method ( judgment matrix ), determine the influence of each layer of indicators on the previous layer of indicators, and linearly weight to obtain the total evaluation target value.
plan
- decision-making level
- According to the judgment matrix, find the relative weight
- Consistency check
- Just pass
Insert image description here
ps: Problems with
judgment matrix AHP
2. Network Hierarchy Analysis Method
The English term for network analysis is Analytic Network Process, or ANP for short. It is a decision-making method proposed by American Professor Saaty in 1996. This method is based on the analytic hierarchy process and is a method that adapts to the non-independent hierarchical structure. . Compared with AHP, ANP replaces the hierarchical structure with a network structure, takes into account the correlation between elements, replaces the linear hierarchical structure with a non-linear structure, adds a feedback mechanism, and takes into account the influence of low-level elements on high-level elements. Dominance.
plan:
analyse problem
- analyse problem
- Construct control layer, network layer structure
- Compare pairwise to construct the influence judgment matrix (1 or 0) of all elements Cj, that is, a supermatrix
- Pairwise comparison, importance matrix of each element
- Determine the weight of each element group of the supermatrix
- Compute weighted supermatrix
The difference between the two:
ANP takes into account the interaction between each element group and the elements
This article is derived from Tsinghua University courseware - Analytical Hierarchy Process AHP and Network Analysis ANP
3. Fuzzy analytic hierarchy process FAHP
The importance matrix changed from 1 to 9 to 0.1 to 0.9. 0.1 to 0.5 is from extremely unimportant to equally important, and 0.5 to 0.9 is from equally important to extremely important. The interval difference between the scales is only 0.1, making it difficult to express the relationship between factors. Extremely important.
4. Double base point method TOPSIS
[Basic Principle] Sort by detecting the distance between the evaluation object and the optimal solution and the worst solution. If the evaluation object is closest to the optimal solution and at the same time far away from the worst solution, it is the best.
[Disadvantages] A certain element may just happen to be similar to the target curve, but not necessarily related. Not necessarily a causal relationship.