Professional Documents
Culture Documents
1. OBJECTIVE:
To conduct comparative analysis of facilities based on Analytical Hierarchical Process (AHP)
using Expert Choice for enhancing judgment of decision process to select optimal location of the
following facilities:
Note: Every student in a group will be assigned a different facility from the above facilities to
conduct this experiment.
2. APPARATUS:
Personal computer,
Expert Choice Software
3. THEORY:
The analytic hierarchy process (AHP), developed by Thomas L. Saaty, is designed to solve
complex multi criteria decision problems. AHP requires the decision maker to provide judgments
about the relative importance of each criterion and then specify a preference for each decision
alternative using each criterion. The output of AHP is a prioritized ranking of the decision
alternatives based on the overall preferences expressed by the decision maker.
Consider a car purchasing decision problem. After a preliminary analysis of the makes and
models of several used cars, list of decision alternatives for three cars is, Honda Accord, a
Saturn, and a Chevrolet Cavalier. Table 1 illustrates the three alternatives of the cars and their
characteristics i.e, factors that can influence on the selection of car.
1. How the four criteria contribute to the overall goal of selecting the best car
2. How the three cars compare using the Price criterion
3. How the three cars compare using the MPG criterion
4. How the three cars compare using the Comfort criterion
5. How the three cars compare using the Style criterion
Pairwise comparisons form the fundamental building blocks of AHP. In establishing the
priorities for the four criteria, AHP will require Diane to state how important each criterion is
relative to each other criterion when the criteria are compared two at a time (pairwise). That is,
with the four criteria (Price, MPG, Comfort, and Style) Diane must make the following pairwise
comparisons:
The scores are given to the factors based on their priorities from the following preference Table
2:
Table 2: Standard preference table
Preference level Score
Equally preferred 1
Equally to moderately 2
Moderately preferred 3
Moderately to strongly preferred 4
Strongly preferred 5
Strongly to very strongly preferred 6
Very strongly preferred 7
Very strongly to extremely preferred 8
Extremely preferred 9
Table 3 illustrates the summary of six pairwise comparisons provided for the car selection
problem.
Table 3 summary of six pairwise comparisons provided for the car selection problem
Pairwise comparison matrix among the factors is constructed once their priorities are known. The
pairwise comparison matrix for the car selection problem is given in Table 4.
The pairwise comparison matrix is filled with the numbers given to their corresponding
priorities as given in Table 5.
Table 5 Pairwise comparison of factors filled with their scores based on their priority.
Because the diagonal elements are comparing each criterion to itself, the diagonal elements of
the pairwise comparison matrix are always equal to 1. All that remains is to complete the entries
for the remaining cells of the matrix. To illustrate how these values are obtained, consider the
numerical rating of 3 for the Price-MPG pairwise comparison. This rating implies that the MPG
Price pairwise comparison should have a rating of 1/3. That is, because Diane already indicated
Price is moderately more important than MPG (a rating of 3), we can infer that a pairwise
comparison of MPG relative to Price should be 1/3. Similarly, because the Comfort-MPG
pairwise comparison has a rating of 4, the MPG-Comfort pairwise comparison would be 1/4. The
complete pairwise comparison matrix for the car selection criteria is given in Table 6.
Table 6 Complete pairwise comparison matrix.
Step 4: Synthesization
Using the pairwise comparison matrix, we can now calculate the priority of each criterion in
terms of its contribution to the overall goal of selecting the best car. This aspect of AHP is
referred to as synthesization. The exact mathematical procedure required to perform
synthesization is beyond the scope of this text. However, the following three-step procedure
provides a good approximation of the synthesization results:
3. Compute the average of the elements in each row of the normalized pairwise comparison
matrix; these averages provide the priorities for the criteria.
In this step pairwise method is used to determine the priorities for the three cars using each of the
criteria: Price, MPG, Comfort, and Style. Determining these priorities requires pairwise
comparison preferences for the cars using each criterion one at a time. For example, using the
Price criterion, pairwise comparisons of the cars is required
Table 6 shows the summary of the car pairwise comparisons provided for each criterion of the
car selection problem. Using this table and referring to selected pairwise comparison entries,
Suppose following preferences are given
In terms of Price, the Cavalier is moderately to strongly more preferred than the Accord.
In terms of MPG, the Cavalier is moderately more preferred than the Saturn.
In terms of Comfort, the Accord is very strongly to extremely more preferred than the
Cavalier.
In terms of Style, the Saturn is moderately more preferred than the Accord.
A synthesization is conducted for each pairwise comparison matrix, using the three-step
procedure described previously for the criteria pairwise comparison matrix.
Four synthesization computations provide the four sets of priorities shown in Table 7. Using this
table, we see that the Cavalier is the preferred alternative based on Price (0.557), the cavalier is
the preferred alternative based on MPG (0.639), the Accord is the preferred alternative based on
Comfort (0.593), and the Saturn is the preferred alternative based on Style (0.656). At this point,
no car is the clear, overall best. The next section shows how to combine the priorities for the
criteria and the priorities in Table 6 to develop an overall priority ranking for the three cars.
In steps 2, 3 & 4 we used Diane’s pairwise comparisons of the four criteria to develop the
priorities of 0.398 for Price, 0.085 for MPG, 0.218 for Comfort, and 0.299 for Style. Now we
want to use these priorities and the priorities shown in Table 4 to develop an overall priority
ranking for the three cars. The procedure used to compute the overall priority is to weight each
car’s priority shown in Table 4 by the corresponding criterion priority.
Multiply by
Overall Priority of the Accord:
Repeating this calculation for the Saturn and the Cavalier, we obtain the following results:
Overall Priority of the Saturn:
Ranking these priorities, we have the AHP ranking of the decision alternatives:
There are three alternatives given to you in Figure 1 for building one of the following facilities
Ware house facility of a product
Hospital facility
Bank Facility
Table 5. Information for the Ware house facility location selection Problem
Factor Alternative 1 Alternative 2 Alternative 3
A1 A2 A3
Land cost F1 12000 $ 15000$ 14000$
Type of land F2 Moderately preferred Equally to
over A2 moderately
over A2
Availability F3 Strongly preferred Moderately
manpower and skills over A1 preferred
Over A1
Cost of labor F4 Strongly to very Equally to
strongly preferred moderately preferred
Over A3 Over A3
Housing availability F5 Strongly preferred Equally to
for employees Over A3 moderately
over A1
Communication F6 Very strongly Moderately preferred
preferred Over A3
Over A2
Raw material F7 Equally to moderately Strongly to very
availability preferred strongly preferred
Over A3 Over A3
Availability of F8 Equally to moderately Strongly preferred
Vendors over A2 Over A3
Maintenance F9 Equally to moderately Very strongly
preferred preferred
Over A2 Over A1
Highway and roads F10 Moderately preferred Moderately
over A2 preferred
over A2
Airport availability F11 Equally to Moderately
moderately over A1 preferred
over A2
Safety F12 Moderately preferred Equally to
over A3 moderately preferred
Over A3
Taxation F13 Equally to moderately Very strongly
over A3 preferred
Over A3
Cultural influence F14 Moderately preferred Moderately
over A2 preferred
over A1
Table 6. Information for the Hospital facility location selection Problem:
b) Install the Expert Choice software from Application file in unzipped folder.
To create your model, “Choosing a site for Facility Location”, perform the following steps:
1. Using the Welcome to Expert Choice dialogue box, select the Create new model radio
button and click OK.
Note: To open an existing model, select Recent; then select a model from the drop-down list or
select Browse existing files.
2. Type the name of your new model such as “My Facility” and select Open. If want to select a
different directory navigate to the directory before entering the model name.
Next you must define your goal.
3. Enter a description of your goal by typing: Choose a site for Facility Location.
4. If you made a typo, select Edit, Edit node, or alternatively, right-click on the Goal and
click Edit node. Then fix your error and press Enter.
Note: A node is defined as an element in the hierarchy that includes a goal, objectives and
alternatives.
5. At this point, you will have a model with only the Goal shown in the Tree View pane. Now
you will add objectives and alternatives to the Facility model.
6. Enter Objectives or Criteria as follow:
Select Edit; then select Insert Child of the Current Node.
Type over the highlighted text and enter the first objective, Cost of Land . When done
press Enter.
The next node will be displayed; type Availability of Raw Material Press Enter.
Repeat the last two steps to enter all criteria.
After entering last criteria a blank node will appear, press Enter or Esc to end the
insertion process.
7. Enter Alternatives as follow:
Select Edit; then select Alternative, Insert and type the alternative name, Mirpur
Mathelo
Press Enter. Alternatively right-click in the alternative pane and select Insert.
Repeat the above step to enter other two alternatives. (Try clicking the Add
Alternative button.)
Your model is now complete as in Figure 5. Save it by selecting: File, Save.
Figure 5. Sample My Facility Model View
Step 3: Make Pairwise comparison
Fill the following table of Pairwise Comparison Matrix using values from relevant table in
Tables 5 to Table 7.
F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14
F1
F2
F3
F4
F5
F6
F7
F8
F9
F10
F11
F12
F13
F14
1. Click on goal in model view and then click on Pairwise Numerical comparison tab to carry
out the comparison. Enter value in each cell comparison matrix, as you mentioned in above
table, with the help of judgment scale.
Priority
F1
F2
F3
F4
F5
F6
F7
F8
F9
F10
F11
F12
F13
F14
Pairwise comparison of alternatives
3. Select an alternative and Click on assessment and go to ‘Pairwise’ or Press ‘Ctrl+P’ and
conduct comparison of all alternative’s with respect to each objective.
Fill the priority of criteria wise each alternative in the following table using values from software
5. Results
Best Alternative
To examine the synthesis:
1. Select Synthesize tab or click while selecting Goal to produce the display of
results.
2. To view the details, click the Details tab.
3. Select Show Totals, By Alternatives to see the breakdown of priorities.
Fill the following table to write down the priority of each alternative using software results
Alternative Priority
A1
A2
A3
EXPERIMENT DOMAIN:
Note: Every student in a group will be assigned a different facility from the above facilities to
conduct this experiment
Objective:
Apparatus:
Procedure:
Note: Use following tables in writing your procedure.
Pairwise comparison of factors
Fill the following table of Pairwise Comparison Matrix using values from relevant table in
Tables 5 to Table 7.
F1 F2 F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14
F1
F2
F3
F4
F5
F6
F7
F8
F9
F10
F11
F12
F13
F14
Fill the priority of each factor in the following table using values from Analysis of software.
Priority
F1
F2
F3
F4
F5
F6
F7
F8
F9
F10
F11
F12
F13
F14
Fill the priority of criteria wise each alternative in the following table using values from software
7. Comments:
………………………………………………………………………………………………
………………………………………………………………………………………………
………………………………………………………………………………………………
………………………………………………………………………………………………
………………………………………………………………………………………………
Taxonomy Level P1 P2 P2 P2 A1 A2 C3
Marks distribution 3 5 3 3 2 2 2
Marks Obtained