Dissertation - Neuro Fuzzy Model

DISSERTATION

TABLE OF CONTENT
DISSERTATION.................................................................................................................................1
CHAPTER – 1......................................................................................................................................3
INTRODUCTION................................................................................................................................3
1.1 Overview of the study............................................................................................................3
1.2 Statement of the problem.......................................................................................................4
1.3 Significance of the study........................................................................................................4
1.4 Research aims and objectives................................................................................................5
1.5 Research questions.................................................................................................................5
1.6 Structure of the dissertation...................................................................................................5
CHAPTER 2.........................................................................................................................................7
LITERATURE REVIEW.....................................................................................................................7
CHAPTER 3.......................................................................................................................................30
RESEARCH METHODOLOGY.......................................................................................................30
3.1 Introduction..........................................................................................................................30
3.2 Research approach...............................................................................................................30
3.3 Research philosophy............................................................................................................30
3.4 Research type.......................................................................................................................31
3.5 Data collection.....................................................................................................................32
3.6 Sampling..............................................................................................................................33
3.7 Data analysis........................................................................................................................33

3.8 Ethical consideration............................................................................................................34
3.9 Research Limitations...........................................................................................................34
3.10 Validity and Reliability......................................................................................................35
CHAPTER 4.......................................................................................................................................37
Data analysis.......................................................................................................................................37
Chapter 5............................................................................................................................................51
Conclusion and Recommendation......................................................................................................51
REFERENCES...................................................................................................................................54

CHAPTER – 1
INTRODUCTION
1.1 Overview of the study
Now-a-days, in the advancement of technology, software developer designs various
software’s that perform specific tasks to automate the transactions or simplify the complex
manual procedures. There are number of activities that are performed or executed by the
software developer such as programming, coding, testing and many others. It requires extensive
efforts to design the software, still developers face challenging conditions in accurate cost
prediction. In the modern period, enormous methods exists to the software design to measure
their cost mainly categorized into two, algorithm as well as non-algorithm that exhibits the
relationship between software size and efforts. It shows that complexity of the software
development procedure requires huge efforts of the developer and incurs high cost. During the
early stage of project development, there are no enough set or quantum of data available.
Moreover, the information that is available is not always consistent & uncertain also; therefore,
developers cannot measure or estimate the cost reliably and accurately.
Now-a-days, with the increase in number of software’s, its sizes and significance, it
becomes really very important for the developer to accurately determine the cost of the software.
In such respect, the process of cost estimation is regarded as parametric estimation that indicates
the relationship between various discrete parameters and the expected outcome of the desired
software. There are various problems that has been encountered or discovered as hurdles in the
cost estimation process such as non-linear relationship with the output, incomplete or imprecise
set of data, complexities in the algorithm and non-algorithm models. Neuro-Fuzzy model also
has been developed to overcome or combat such issues through a combination of both the
artificial neural network & Fuzzy logic as well that produce reliable and right figure regarding
software cost.
As said earlier, that programmers allocate time on various activities to design the
software, however, its cost estimation tends to be hard because of lack of sufficient information
or dataset. On the other hand, if cost can be estimated reliably, then still, software developer
spend a lot of time and put great efforts for such estimation that is ignored by the programmers in
cost estimation. Apart from this, there are number of changes required till the end of final
software design that is available for use, the cost of these changes are also not considered by the

developers to accurately determine their software cost. Previous research studies discovered that
connectivity, networking and communication failure, hardware and software cost, lack of right
measurement of efforts contributed and others are exists to the programmers which mislead the
cost measurement decisions.
1.2 Statement of the problem
The main concern or problem for carrying out the present investigation is it is very
difficult for the software designer to measure the right cost of their software. However, it is
extremely essential or important requirement for the developers to measure the cost of their
designed software. Although there are wide ranges of methods that are incorporated by the
designers to measure their cost like algorithm and non-algorithm based methods, still it is too
difficult and complex for the designers to determine the right cost of their newly developed
software’s. There are various difficulties faced by the developers in cost estimation like cost of
testing, technical or network failure, cost of efforts and others are the main problems that are
often faced by the developers in accurately determine the cost of the software. There are very
few researches that had been conducted earlier to examine various fundamentals of cost
measurement in software development, still, none of the researcher discovered all the factors,
components or reasons that mislead the cost measurement decisions. Therefore, the study will
bridge the gap through discovering the main challenges and difficulties in estimating or
determining the reliable or accurate cost of new software development.
1.3 Significance of the study
As already said, that there are very few researches that had been conducted earlier by the
scholars in the current area. Therefore, the current study will provide great assistance to those
investigators who are interested in carrying an investigation in the chosen area of study. It is
because; this study will be work as an basis that provides an in-depth knowledge, conceptual and
theoretical understanding and methodological contribution of the topic. Thus, with the help of
this, scholar will be able to easily and clearly understand the topic and carry out an investigation
in the related field or area. Moreover, software developers or programmers will also get benefits
from the present research through identifying various existing cost measuring models, difficulties
and thereby select the most appropriate method for finding the cost of their constructed software.
With this, they can also make smarter pricing decisions for the customers to get an adequate

return on their software. The main aim of the study behind conducting the present investigation
is to discover or identify the factors that cause issues for the programmers in development
1.4 Research aims and objectives
Research aim: To analyze the factors that cause issues in an accurate cost estimation of a new
developed software
Research objectives: The main target of the study will be achieved through the following framed
objectives that are stated here as under:
 To investigate the fundamentals of cost estimation of a software project.
 To examine various models available to the software developers to measure their cost of
the software.
 To discover challenges or issues that is faced by the software developer in accurate cost
estimation.
 To recommend the best model for measuring accurate or reliable cost of the new
designed software.
1.5 Research questions
RQ:1. What are the basic fundamentals of cost measurement of a newly designed software?
RQ:2. What kind of methods and techniques are currently used by the programmers to estimate
their software cost?
RQ:3. What are the challenges or difficulties faced by the programmers in determining the
reliable or accurate cost of the software?
RQ:4. Which is the best model or method that can be used by the software developers to measure
an accurate cost of the designed software?
1.6 Structure of the dissertation
This section is referred to be the blue print of the study where it has simply provided an
overall background of this thesis. It is basically to build a stepwise approach for the readers to
make a primary interpretation of this structure to easily go through the factual content of the
current thesis.
Chapter 1 Introduction: It is the foremost part of this research work that has clearly outline the
overall background of the study with a precise identification of its aims, objectives, significance
and specifications, etc.
Chapter 2 Literature Review: This section has represented to be yet another conceptual part of
this investigation that has critically discussed upon distinct perception of authors as a way to

determine the problematic factors associated with the consent of cost management of software
development. It is basically done by determining the projected objectives of the study and
framing accordant themes in sub-parts to assist the researcher in assessing the critical standpoints
of varied authors.
Chapter 3 Research Methodologies: This unit has revealed the tact of carrying a refined study
with some implicated tools and techniques of research. This part has thus discoursed about the
undertaken methods of investigation by the surveyor.
Chapter 4 Data analysis: It is yet another significant portion that has deliberated the obtained
facts and findings of this research work where a qualitative method of analysis has been used by
the investigator to interrogate into the recent subject matter.
Chapter 5 Conclusion and recommendation: It is the eventual section of this investigation that
has concluded the entire study in relevance to the above discussed sub divisions. It is basically to
confirm the achievement of undertaken goals and objectives by together recommending certain
effective ways to reach to the targeted agenda of the survey.

CHAPTER 2
LITERATURE REVIEW
Introduction
Software development projects are very often characterized by severe cost
overruns, schedule slippages and poor quality of deliverables. In order to meet
the quality expectations of customers, lot of time and effort is spent by software
seller organizations on fire fighting after the software product is released
(attending to failures/ fixing critical bugs etc.) which has a cost implication.
Unfortunately, these costs are not clearly understood. These costs often
disappear as the costs of testing, the general developments costs, or the
operating expenses which is misleading.
The basic principle of software process management is that if the development
process is under statistical control, a consistently better result can only be
produced by improving the process. If the process is not under statistical