Scribd
Upload
403 views10 pages

ANP

The analytic network process (ANP) is a decision-making method that involves 6 steps: 1) Define a network structure of criteria and alternatives. 2) Identify dependencies between components. 3) Create pairwise comparison matrices through surveys. 4) Fill out a supermatrix with component weights. 5) Calculate a weighted supermatrix by normalizing columns. 6) Raise the weighted supermatrix to higher powers to converge on a stable limit matrix for making decisions.

Uploaded by

Roumaissa Gueddouche
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
Download as pptx, pdf, or txt
403 views10 pages

ANP

The analytic network process (ANP) is a decision-making method that involves 6 steps: 1) Define a network structure of criteria and alternatives. 2) Identify dependencies between components. 3) Create pairwise comparison matrices through surveys. 4) Fill out a supermatrix with component weights. 5) Calculate a weighted supermatrix by normalizing columns. 6) Raise the weighted supermatrix to higher powers to converge on a stable limit matrix for making decisions.

Uploaded by

Roumaissa Gueddouche
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1/ 10

Analytic network

process

Presented by: gueddouche roumaissa


ANP Steps
- Step 1: Based on the decision goal, a network structure
including criteria and alternatives clusters should be
configured.

- Step 2: The dependences among all components of the


previous structure should be identified and listed, in order
that the impacts among them can be defined.
• Step 3: Pairwise comparison matrices of the components
with IC < 10%

• Step 4: Supermatrix and Cluster Matrix have to be filled with


the weights elicitated during Step 3
• Step 5: Weighed Supermatrix calculation. To compute the
Weighted Supermatrix, the Supermatrix is transformed to
column stochastic, If needed the columns are again
normalized to keep summing to 1.

• Step 6: The Weighted Supermatrix should be limited by raising


it to a sufficiently large power until it converges into a stable
limit matrix.
example
• Probleme:
choosing phone from three options Model A,B or C
considering three criteria touch screen ,camera and battery
Step1:defining clusters
• Step2: relationship matrix
ALTERNATIVE CEITERIA

Model A Model B Model C Touch battery camera


screen

ALTERNATIVE Model A 0 0 0 1 1 1

Model B 0 0 0 1 1 1

Model C 0 0 0 1 1 1

CRITERIA Touch 1 1 1 0 0 0
Screen
Battery 1 1 1 0 0 0

camera 1 1 1 0 0 0
• Step 3: Pairwise comparison matrices of
the components

Alternatives
comparison matrix

battery Model Model B Model C weight


A

Model A 0,188 0,273 0,623


0,179
We do the same
Model B 0,750 0,636 0,137 for the rest of
0,718 the criterions
Model C 0,063 0,091 0,239
0,103
• step 4: Supermatrix
Model A Model B Model C Touche battery camera
screen
Model A 0 0 0 0,093 0,625 0,627

model B 0 0 0 0,292 0,137 0,094

0 0 0 0,615 0,238 0,28


Model C
touch 0,211 0,211 0,211 0 0 0
screen
Battery 0,705 0,705 0,705 0 0 0

camera 0,084 0,084 0,084 0 0 0

step 5 :we normalize the supermatrix to get


The Weighted Supermatrix
• Step 6: limit matrix
Model A Model B Model C Touch Battery Camera
screen
Model A 0,256 0,256 0,256 0,256 0,256 0,256

Model B 0,083 0,083 0,083 0,083 0,083 0,083

Model C 0,161 0,161 0,161 0,161 0,161 0,161

Touch 0,105 0,105 0,105 0,105 0,105 0,105


screen
Battery 0,352 0,352 0,352 0,352 0,352 0,352

Camera 0,042 0,042 0,042 0,042 0,042 0,042


Software ANP SOLVER

You might also like