Fuzzy Comprehensive Evaluation -----Analytic Hierarchy Process AHP

Fuzzy Comprehensive Evaluation -----Analytic Hierarchy Process AHP

The text reads first:

​ Recently, there is a topic that needs to use fuzzy comprehensive evaluation. The author also collected information from various parties, and finally found an article on XX. It is simple, systematic, and detailed, and there is no shortage of easy-to-understand and in-depth insights. Here is a link , readers can try to read by themselves, and there will be a lot of gains.

​ Later, when it was discovered that the AHP was needed, the big brother took it all in one stroke, but as newcomers, we need to find information. The author came here uninvited, and made my humble opinion on the AHP, hoping to learn from each other.

---

1. research questions
2. construct hierarchy
3. Build a judgment matrix
4. Calculate the corresponding weight of each layer element
5. get conclusion

Introduction: Give the weight a little scientifically

1. Questions to be studied

Topic: Xiao Ming wants to travel. There are three alternative destinations: Nanjing, Guilin and Sanya . The main considerations are the landscape, food and accommodation, culture and price of these destinations . Now it is necessary to scientifically select Xiao Ming based on each element The final destination.

2. Build a hierarchy

3. Build a judgment matrix

4. Calculate the corresponding weight of each layer element

4.1 Normalize by column, mean by row

4.2 Make a consistency judgment on the matrix

Judge whether the given score is reliable or not, and avoid inconsistency in the score.

If CR<0.1, it means the test is passed.

4.3 For each index value, for all corresponding programs, the weight of each program

There are four index values ​​in this example, namely A1 scenery, A2 food and accommodation, A3 price, A4 humanities. This section uses scenery as an example to explain

After that, normalize by column -> mean by row

Consistency checks are then performed.

repeat~

4.4 Get the total judgment matrix

Among them, the red box is the w calculated in 4.1 (4.1 is an approximate value), the blue box is the w calculated in 4.3, and the number in the green box is obtained by multiplying and adding the corresponding data in the second column and the first column value.

Consistency check for the final conclusion

5 results

The color box is the w calculated in 4.3, and the number in the green box is the value obtained by multiplying and adding the corresponding data in the second column and the first column.

Consistency check for the final conclusion

5 results

The final result (0.56, 0.25, 0.19), the final destination is Nanjing.