You are on page 1of 15

Module : Supply Chain Modelling And Analytics

(BUSN8034)

Assessment 3 – Individual Final Report

Student Name : Arjun Raveendranath


Login : ar797
Date : 10/05/2023

1|Pag e
Table Of Contents
1. Problem 1……………………………………………………………………………………………………3
1.1. Multiple Regression Model…………………………………………………………….3
2. Problem 2……………………………………………………………………………………………………6
3. Problem 3……………………………………………………………………………………………………8
4. Appendix A: AON Network……..……………………………….………………………………….13
5. Appendix B: Regression dialogue box………………………………………………………...13
6. Appendix C: Output of case 1…………………………………………………………………….14
7. Appendix D: Output of case 2…………………………………………………………………….14
8. Appendix E: Network information and activity times………………………………….15
9. Appendix F: Jake’s pair-wise comparison inputs of the four criteria……………15

2|Pag e
1. Problem 1

The aim of this report is to predict the monthly energy use of Skytech Builders. The company’s
management has analyzed that their monthly energy use is a function of production volumes, daily
outside temperature, and number of working days. In this case, a multiple regression model has been
built using Excel with the data collected over the last year and used to predict the monthly energy use.
There are 3 independent variables namely production volumes, daily outside temperature and number
of working days and 1 dependent variable which is the energy use.

1.1 Multiple Regression Model

To create a multiple regression model in excel, the Analysis ToolPak was load using the following steps:

 Select File -> Options -> Add-Ins


 Select “Excel Add-ins” and click “Go”
 Check the “Analysis ToolPak” and hit “OK”.

Once it is loaded, click on “Data Analysis” under the “Data” tab. Then choose “Regression” and click
“OK” and the box displayed below will appear.

Figure 1: Regression dialogue box

In the above box, choose the data of the dependent variable “Energy Use” in “Input Y Range” and the
independent variables “production volumes”, “daily outside temperature” and “number of working
days” in “Input X Range” and choose a cell for the output range. The confidence level has been set at
95% since we are testing the hypothesis for that significance or there is relationship between x and y,
where x is the independent variables and y is the dependent variable. Then click “OK”.

The regression model has been created with all the details and calculations as summary as given
below:

3|Pag e
Figure 2: Output of case 1

In the above summary, it shows that the:

 Significance F as 0.004134161 which is less than 0.05 which shows that or there is a
relationship between x and y.
 R Square has a value of 0.79290153 or close to 80% which is a good value as R Square shows
the goodness or fit of the hypothesis. It is also to be noted that as more variables are added
to the hypothesis, the R square value increases.
 Adjusted R Square Value is 0.7152396. It is also to be noted that as more variables are added
to the hypothesis, the Adjusted R Square value increases which indicates that the accuracy of
the model also increases.
 The P-values for the intercept, x variable 1, x variable 2 and x variable 3 are 0.167275626,
0.009346249, 0.300503737, and 0.044773784 where the P-values for intercept and x variable
2 are greater than 0.05 and hence this hypothesis is rejected.

Now, the same procedure is repeated using “Data Analysis” under the data tab without the
independent variable “number of working days” and the result obtained is as given below:

4|Pag e
Figure 3: Output of case 2

In the above summary, it shows that the:

 Significance F as 0.001589614 which is less than 0.05 which shows that or there is a
relationship between x and y.
 R Square has a value of 0.761183231 or close to 80% which is a good value as R Square shows
the goodness or fit of the hypothesis.
 Adjusted R Square Value is 0.708112837.
 The P-values for the intercept, x variable 1, and x variable 2 are 0.010532, 0.000743, 0.035759.
Since all the values are below 0.05, this model can be accepted.

Though the R Square and the Adjusted R Square value has reduced, there is only very small variation
which does not affect the accuracy of the model much.

Generally, for a given set of outputs Y and a set of n input values X1, X2,….,Xn, the aim is to determine
the parameters b0, b1,…., bn of the the linear equation:

y = b0 + b1x1 + …… +bnxn

Therefore, from the above summary, we can find that the linear equation for this model will be

y = 229.89 + 1.33 x1 + 1.37 x2

where,

x1 is the temperature

x2 is the production

(i) We do not need to include all the variables in this model as we are not considering the independent
variable “number of workdays”.

(iii) The predicted energy use for a month of 25 days, temperature 55 degrees and production 120
units would be

5|Pag e
y = 229.89 + 1.33*55 + 1.37*120

y = 467.44

Therefore, the predicted energy use would be 467.44.

2. Problem 2

This study examines the project network and related activity periods for the most recent client
engagement of International Capital Inc. A small-to-medium-sized logistics company may turn to IC, an
investment bank, for funding to set up its facilities. The critical path and slack periods are calculated
when the project network is built in the report. The activities along the critical path are those that
cannot be postponed without causing the project as a whole to be delayed. The report also examines
the likelihood of finishing the project within a specified time range as well as the project's overall
duration, earliest and latest start and conclusion times.

Activity Description Immediate Predecessor


A Research Client Firm --
B Create due diligence rough report A
C Coordinate needs proposal with B
client
D Estimate future demand and cash B
flows
E Draft future plans for client D
company
F Create and approve legal B
documents
G Integrate all drafts into the first- C, E, F
draft proposal
H Line up potential sources of capital G
I Check, approve and print the final G
legal proposal
J Sign contracts and transfer funds H, I
Figure 4: Network information and activity times

With the information given in the table above an AON or Activity-On-Node network has been created
to find the critical path. An AON network is a network diagram in which the nodes represent the various
activities and the arrows show the logical connections between them.

6|Pag e
C H

Start E G Finish
A B D J

F I

Figure 5: AON Network

After creating an AON or Activity-On-Node network, we have decided to do PERT or Program


Evaluation Review Technique analysis to find the critical path which will be the expected time to
complete this project. PERT or Program Evaluation Review Technique analysis is used to calculate the
bare minimum time needed to finish a project, it examines the length of time needed to finish each
task and the dependencies that go along with it.

In a PERT analysis, to get the probabilistic time estimates, each activity time follows a Beta – PERT
distribution.

Mean, te = (a + 4m + b)/6

where,

a is the optimistic finish time

m is the most likely finish time

b is the pessimistic finish time

Variance, σ2 = (b - a / 6)2

The PERT analysis with the given data is as given below:

Activity Optimistic, a Most Likely, m Pessimistic, b Mean,te Variance


A 4 7 10 7 1
B 2 5 8 5 1
C 16 19 28 20 4
D 6 9 24 11 9
E 1 7 13 7 4
F 4 10 28 12 16
G 2 5 14 6 4
H 5 8 17 9 4
I 2 5 8 5 1
J 17 29 47 30 25

7|Pag e
So, from the AON network and the PERT analysis, we could find all the possible paths and the expected
completion time for each paths as mentioned:

1. A->B->C->G->H->J - 77 Days
2. A->B->C->G->I->J - 73 Days
3. A->B->D->E->G->H->J - 75 Days
4. A->B->D->E->G->I->J - 71 Days
5. A->B->F->G->H->J - 69 Days
6. A->B->F->G->I->J - 65 Days
Therefore, the mean would be the sum of the means in the critical path which is 77 and the variance
would be the sum of the variances in the critical path is 39.

(i) Since the path A->B->C->G->H->J has the longest number of days which is 77 days, it has to be
considered as the critical path.

(ii) The probability of completing the project in 85 days or less would be

P(T<=85) = NORMDIST (T, Mean, Std Dev, 1)

= NORMDIST (85, 77, SQRT(39), 1)

= 0.8999076 (or) 90%

(iii) The probability that the project would exceed 90 days would be 1 - P(T<=90), i.e.

P(T<=90) = NORMDIST (T, Mean, Std Dev, 1)

= NORMDIST (90, 77, SQRT(39), 1)

= 0.98134

Therefore, 1 - P(T<=90) = 1 - P(T<=90)

= 0.01868649 (or) 1.90%

3. Problem 3

The aim of this report is to help Jake buy a used car. He has narrowed his options down to three
vehicles: the Honda Accord, the Saturn, and the Chevrolet Cavalier. For the purpose of buying an
automobile, Jake has determined that four factors are important: price, miles per gallon (MPG),
comfort, and style. Jake has given his pairwise comparison inputs for the four criteria and also pairwise
comparison matrices for each car using each criterion. We will be using AHP or Analytical Hierarchy
Process to help Jake make the best decision. The Analytic Hierarchy Process (AHP) is a method created
to quantify managerial assessments of the relative weight given to each of the many competing factors
considered during the decision-making process. The strategy is well-liked and frequently used in multi-
criteria decision-making. It enables the use of qualitative and quantitative evaluation criteria.

AHP or Analytical Hierarchy Process for the above problem with the given inputs is done as follows:

8|Pag e
Figure 6: Jake’s pair-wise comparison inputs of the four criteria

Step 1 – Pairwise Comparison Matrix

 We need to create a pairwise comparison matrix using the given pairwise comparison inputs
of the criterias.
 In the matrix, the diagonals will have the number 1 as it is comparing the same criterion.
 Enter the inputs as given according to its importance.
 In the remaining cells, the input will be the reciprocal of its corresponding cell.

The pairwise comparison matrix was found to be as shown below:

Price MPG Comfort Style


Price 1.00 3.00 2.00 2.00
MPG 0.33 1.00 0.25 0.25
Comfort 0.50 4.00 1.00 0.50
Style 0.50 4.00 2.00 1.00

Step 2 – Develop a normalized matrix

 To develop a normalized matrix, we need to divide each number in a column of the pairwise
comparison matrix with the sum of its column.
 The resulting normalized matrix must have the sum of the elements of each column to be 1.

The resulting normalized matrix was found to be as given below:

Normalised Matrix
0.43 0.25 0.38 0.53
0.14 0.08 0.05 0.07
0.21 0.33 0.19 0.13
0.21 0.33 0.38 0.27

Step 3 – Develop the priority vector

 Calculate the weighted average by getting the average of each row in the normalized matrix.

9|Pag e
 The weighted average will be in the form of a vector and is denoted by w1, w2, w3 and w4.
 The sum of all the values in the vector must be equal to 1.

The resulting vector was found to be as given below:

Weights
0.40
0.09
0.22
0.30

Step 4 – Determine the score of each alternative on each objective

 Repeat step 2 and step 3 for Jake’s other pair-wise comparison matrix for each of the cars
using each criteria.
 Find the weighted average or the scores for all the cars based on each criteria and their
weights.

The scores for all the cars based on each criterion is calculated as given below:

PRICE

Price Accord Saturn Chevy Normalised Matrix


Accord 1.00 0.33 0.25 0.13 0.10 0.14
Saturn 3.00 1.00 0.50 0.38 0.30 0.29
Chevy 4.00 2.00 1.00 0.50 0.60 0.57

Scores - Price
0.12
0.32
0.56

MPG (or) Miles Per Gallon

MPG Accord Saturn Chevy Normalised Matrix


Accord 1.00 0.25 0.17 0.09 0.06 0.11
Saturn 4.00 1.00 0.33 0.36 0.24 0.22
Chevy 6.00 3.00 1.00 0.55 0.71 0.67

Scores - MPG
0.09
0.27
0.64

10 | P a g e
COMFORT

Comfort Accord Saturn Chevy Normalised Matrix


Accord 1.00 2.00 8.00 0.62 0.63 0.53
Saturn 0.50 1.00 6.00 0.31 0.32 0.40
Chevy 0.13 0.17 1.00 0.08 0.05 0.07

Scores - Comfort
0.59
0.34
0.07

STYLE

Style Accord Saturn Chevy Normalised Matrix


Accord 1.00 0.33 4.00 0.24 0.23 0.33
Saturn 3.00 1.00 7.00 0.71 0.68 0.58
Chevy 0.25 0.14 1.00 0.06 0.10 0.08

Scores - Style
0.26
0.66
0.08

Step 5 – Determine the best alternative

To determine the best alternative we need to multiply the weighted average of the pairwise
comparison matrix and the scores of all the criterions of each cars in the form of a matrix as shown
below:

Scores - Scores - Scores - Scores - Weighted


Price MPG Comfort Style Weights Score
Accord 0.12 0.09 0.59 0.26 0.40 0.26
Saturn 0.32 0.27 0.34 0.66 0.09 0.42
Chevy 0.56 0.64 0.07 0.08 0.22 0.31
0.30

So, from the analysis of Analytical Hierarchy Process or AHP, Jake must consider buying the car Saturn
as it has the highest weighted score of 0.42.

11 | P a g e
Step 6 – Check for consistency

We are checking the consistency to check if the output found is accurate or not. Ideally, the consistency
ratio must be less than 0.1 for the result to be acceptable.

Consistency Ratio, CR = CI/RI

where,

CI – Consistency Index

RI – Random Index

 Make a weighted sum of the entries and the vector of weights for each row of the pair-wise
comparison matrix.

To check the consistency of matrix A.

Weights
Aw = 1.00 3.00 2.00 2.00 0.40 1.69
0.33 1.00 0.25 0.25 0.09 0.35
0.50 4.00 1.00 0.50 0.22 0.91
0.50 4.00 2.00 1.00 0.30 1.27

 Calculate the average of the ratios between each element of the weighted matrix and the
corresponding weights in the vector of weights.
((1.69/0.40)+(0.35/0.09)+(0.91/0.22)+(1.27/0.30))/4 = 4.18
 To compute the consistency index,
CI = [(Step 2 result)-n]/(n-1), n = objectives
CI = (4.18 – 4)/3 = 0.06
 To determine the random index, we need to consider the number of objectives which is 4. For
a problem with n = 4, its RI would be 0.9.
 Therefore, the consistency ratio would be:

Consistency Ratio, CR = CI/RI

= 0.06/0.9

= 0.07<1

Since, the consistency ratio is less than 1, the found result is considered to be consistent. Therefore,
Jake can buy the car Honda Saturn without having any second thoughts.

12 | P a g e
Appendix A: AON Network

C H

Start D E G Finish
A B J

F I

Appendix B: Regression dialogue box

13 | P a g e
Appendix C: Output of case 1

Appendix D: Output of case 2

14 | P a g e
Appendix E: Network information and activity times

Activity Description Immediate Predecessor


A Research Client Firm --
B Create due diligence rough report A
C Coordinate needs proposal with B
client
D Estimate future demand and cash B
flows
E Draft future plans for client D
company
F Create and approve legal B
documents
G Integrate all drafts into the first- C, E, F
draft proposal
H Line up potential sources of capital G
I Check, approve and print the final G
legal proposal
J Sign contracts and transfer funds H, I

Appendix F: Jake’s pair-wise comparison inputs of the four criteria

15 | P a g e

You might also like