Scribd
Upload
16 views

Software Engineering Assignment: MOR: Semester 3, 2021-22 - South Campus: Group 3

This software engineering assignment involves comparing two software reliability growth models - the Goel-Okumoto model and the Kapur-Garg model - using a given dataset. The document outlines the assumptions and equations of each model. It then provides the dataset to be analyzed, which contains failure data over 28 weeks of testing. The assignment involves using SPSS to estimate parameters for each model and calculate 12 decision criteria to determine the better fit model for the given data.

Uploaded by

Aamir Iqbal
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
16 views

Software Engineering Assignment: MOR: Semester 3, 2021-22 - South Campus: Group 3

This software engineering assignment involves comparing two software reliability growth models - the Goel-Okumoto model and the Kapur-Garg model - using a given dataset. The document outlines the assumptions and equations of each model. It then provides the dataset to be analyzed, which contains failure data over 28 weeks of testing. The assignment involves using SPSS to estimate parameters for each model and calculate 12 decision criteria to determine the better fit model for the given data.

Uploaded by

Aamir Iqbal
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 9

SOFTWARE ENGINEERING ASSIGNMENT

MOR: Semester 3, 2021-22 | South Campus: Group 3

Submitted to, Submitted by,


Aamir Iqbal
Dr. Adarsh Anand
Medni Aswal
Nehla Shajahan
Pranjol Haldar
Shreya Goel
Vallari Agarwal
To calculate and compare the
12 Decision Criteria or
Indicators of Goodness of Fit
OBJECTIVE for Goel - Okumoto Model
(GO Model) and Kapur and
Garg Model (KG Model) for a
given authentic dataset.
GOEL-OKUMOTO MODEL
Basic assumptions of GO Model are:
1. The failure observation phenomenon is modeled by NHPP
2. Failures are observed during execution caused by remaining faults in the software.
3. All faults are mutually independent
4. The debugging process is perfect and no new fault is introduced during the debugging process i.e. number of
faults in the testing phase are finite.
5. As soon as failure is detected, the corresponding fault is isolated and removed immediately with no additional
error generation and removal.
6. The average failure detection rate depends linearly on the remaining no. of faults i.e.

where,
a- expected no. of faults in the software before testing | b- failure detection rate | m(t)- no. of faults removed unto time t |
[a - m(t)]- is the remaining no. of faults in the software at time t.

Equation of GO model:
KAPUR AND GARG MODEL
On a failure observation, the fault removal phenomenon also removes proportion of remaining faults without
causing any failure.
Basic assumptions of KG Model are:
1. There are no failure at time t=0
2. Software has finite number of faults.
3. The number of faults detected is proportional to current fault content and the proportionality increases
linearly with each additional fault removal
4. Mutually dependent Faults: - which lies on the same execution path of the software.
5. The second type of fault is detectable if the first type of fault has already been detected.
Consider,
m(t) – no. of faults removed by time t | a- expected total no. of faults in the software | p = fault removal rate for
independent faults | q = rate of removing additional faults
Equation of KG model:
Months(t) Failures Months(t) Failures
1 3 15 7
2 3 16 0
3 38 17 2
4 19 18 3

DATASET USED 5
6
12
13
19
20
2
5
7 14 21 2
8 32 22 3
The dataset considered here indicates the
9 8 23 4
complete failure data for a software that
was developed and tested for 28 weeks. 10 8 24 1
11 11 25 2
12 14 26 1
13 7 27 0
14 7 28 1
PARAMETER ESTIMATION USING SPSS
GO- MODEL
KG- MODEL
CRITERIAS FOR GOODNESS OF FIT

You might also like