Fuzzy Based Software Cost Estimation Methods: A Comparative Study

ACM Sigapp Applied Computing Review, 2000

Estimation of effort/cost required for development of software products is inherently associated with uncertainty. In this paper, we are concerned with a fuzzy set-based generalization of the COCOMO model (f-COCOMO). The inputs of the standard COCOMO model include an estimation of project size and an evaluation of other parameters. Rather than using a single number, the software size can be regarded as a fuzzy set (fuzzy number) yielding the cost estimate also in form of a fuzzy set. The paper includes detailed results with this regard by relating fuzzy sets of project size with the fuzzy set of effort. The analysis is carried out for several commonly encountered classes of membership functions (such as triangular and parabolic fuzzy sets). The issue of designer-friendliness of the f-COCOMO model is discussed in detail. Here we emphasize a way of propagation of uncertainty and ensuing visualization of the resulting effort (cost). Furthermore we augment the model by admitting software systems to belong partially to the three main categories (namely embedded, semidetached and organic) and discuss key implications of this generalization and highlight its links with a generalized sensitivity analysis. The experimental part of the study illustrates the approach and contrasts it with the standard numeric version of the COCOMO model.

2007

The development of software has always been characterized by parameters that possess certain level of fuzziness. This requires that some degree of uncertainty be introduced in the models, in order to make the models realistic. Fuzzy logic fares well in this area. Many of the problems of the existing effort estimation models can be solved by incorporating fuzzy logic. Besides, fuzzy logic had been combined with algorithmic, non-algorithmic effort estimation models as well as a combination of them to deal with the inherent uncertainty issues. This paper also described an enhanced fuzzy logic model for the estimation of software development effort. The model (FLECE) possesses similar capabilities as the previous fuzzy logic model. In addition to that, the enhancements done in FLECE improved the empirical accuracy of the previous model in terms of MMRE (Mean Magnitude of Relative Error) and threshold-oriented prediction measure or prediction quality (pred).

Quality Software, 2003. …, 2003

A novel neuro-fuzzy Constructive Cost Model (COCOMO) for software estimation is proposed. The model carries some of the desirable features of the neurofuzzy approach, such as learning ability and good interpretability, while maintaining the merits of the COCOMO model. Unlike the standard neural network approach, this model is easily validated by experts and capable of generalization. In addition, it allows inputs to be continuous-rating values and linguistic values, therefore avoiding the problem of similar projects having different estimated costs. Also presented in this paper is a detailed learning algorithm. The validation, using industry project data, shows that the model greatly improves the estimation accuracy in comparison with the well-known COCOMO model.

A novel neuro-fuzzy Constructive Cost Model (COCOMO) for software estimation is proposed. The model carries some of the desirable features of the neurofuzzy approach, such as learning ability and good interpretability, while maintaining the merits of the COCOMO model. Unlike the standard neural network approach, this model is easily validated by experts and capable of generalization. In addition, it allows inputs to be continuous-rating values and linguistic values, therefore avoiding the problem of similar projects having different estimated costs. Also presented in this paper is a detailed learning algorithm. The validation, using industry project data, shows that the model greatly improves the estimation accuracy in comparison with the well-known COCOMO model.

Software Engineering especially project planning, scheduling, monitoring and control are based on accurate estimate of the cost and effort. In the initial stage of Software Development Life Cycle (SDLC), it is hard to accurately measure software effort that may lead to possibility of project failure. Here, an empirical comparison of existing software cost estimation models based on the techniques used in those models has been elaborated using statistical criteria. On the basis of findings of empirical evaluation of existing models, a Neuro-Fuzzy Software Cost Estimation model has been proposed to hold best practices found in other models and to optimize software cost estimation. Proposed model gives good result as compared to other considered software cost estimation methods for the defined parameters in overall but it is also dependent on type of project, data and technique used in implementation.

As the most uncertain and complex project when compared to other types of projects, software development project is highly depend on the result of software project planning phase that helping project managers by predicting the project demands with respect to the budgeting, scheduling, and the allocation of resources. The two main activities in software project planning are effort estimation and risk assessment which has to be executed together because the accuracy of the effort estimation is highly dependent on the size, nature, and number of project risks, which are inherent in a particular software project.

Journal of Computer Science, 2013

Accurate software development effort estimation is critical to the success of software projects. Although many techniques and algorithmic models have been developed and implemented by practitioners, accurate software development effort prediction is still a challenging endeavor in the field of software engineering, especially in handling uncertain and imprecise inputs and collinear characteristics. In this study, a hybrid intelligent model combining a neural network model integrated with fuzzy model (neuro-fuzzy model) has been used to improve the accuracy of estimating software cost. The performance of the proposed model is assessed by designing and conducting evaluation with published project and industrial data. Results have shown that the proposed model demonstrates the ability of improving the estimation accuracy by 18% based on the Mean Magnitude of Relative Error (MMRE) criterion.

2016

From the very beginning COCOMO model was used for estimation process, at that era fuzzy logic or perhaps artificial intelligence wasn't grounded into solid. In present scenario projects are very large and are generally globally distributed as well as in software improvement, software effort estimation is one of the crucial steps particularly for offshore projects. The principle aim of this paper is always to emphasize all the uncertainties that had been faced in earlier time, but by means of fuzzy logic precision of estimation was improved. Fuzzy-logic technique primarily based software effort estimation models will be more reliable and agreeable, especially for significant and complex initiatives KeywordsEffort Estimation, Fuzzy logic, Constructive Cost model (COCOMO), Fuzzification, Dfuzzyfication.

Over the past 30 years, one of the challenges of software engineers and managers of software companies in the development of software projects is to estimate accurate cost, effort, quality and risk analysis. Although over 30 years several models presented by researchers, each of them had their strengths and weaknesses. But the need for new methods to overcome the COCOMO model still exists. During the last 20 years’ models based on Artificial Intelligence have been considered more than other models by researchers which among them, neuro-fuzzy models are the latest model. The aim of this paper is to reduce errors and increase accuracy in estimating the cost and effort in software development. To achieve these goals NASA63 data collection and Adaptive Neuro Fuzzy Inference System (ANFIS) models are used and we could achieve MMRE error in proposed method to 0/0984 and the accuracy of estimate to 0/889. Keywords: Software Cost Estimation, Neuro-Fuzzy Model, COCOMO Model.

International Journal of Information Technology, 2018

Software cost estimation SCE is directly related to quality of software. The paper presents a hybrid approach that is an amalgamation of algorithmic (parametric models) and non-algorithmic (expert estimation) models. Algorithmic model uses COCOMO II while non algorithmic utilizes Neuro-Fuzzy technique that can be further used to estimate accuracy in irregular functions. For generalization of the model, Neuro-fuzzy membership functions have been used and simulated using mathematical tool MATLAB. Also, the proposed model has been validated with traditional COCOMO model (COCOMO 81) by using NASA software project data. The experimental results suggest that the proposed model gives better SCE as compared to its traditional counterpart.

Financial health of many organizations now-a-days is being affected by investment in software and their cost estimation. Therefore, to provide effective cost estimation models are the most complex activity in software engineering fields. This paper presents a fuzzy clustering and optimization model for software cost estimation. The proposed model uses Pearson product-moment correlation coefficient and one-way ANOVA analysis for selecting several effort adjustment factors. Further, it applies fuzzy C-means clustering algorithm for project clustering. Then, parameters of COCOMO model have been optimized using Multi-objective Genetic Algorithm (MOGA). Here, two objectives are considered. One is to minimize the Mean Magnitude of Relative Error (MMRE) and other is to maximize the Prediction (PRED). This model has been tested on the COCOMO dataset. The optimization result has also been compared with Multi-objective Particle Swarm Optimization (MOPSO) algorithm. The result has proved superiority of MOGA in parameter optimization for getting strength back the accuracy of software cost estimation.

IAEME PUBLICATION, 2019

Effort estimation of software development is a daunting work that is being carried out by software developers as not much of the data about the software which is to be developed is available during the early stages of software development) are fuzzified that leads to reliable and accurate estimation of effort. The results show that the value of Magnitude of Relative Error (MRE) obtained by applying fuzzy logic is quite lower than MRE obtained from algorithmic model. By analyzing the results further it is observed that Gaussian Membership Function performs better than Triangular Membership Function and Trapezoidal Membership Function as the transition from one interval to another is quite smoother. Here varying number of COCOMO II inputs are fuzzified with these membership functions. The main reason for this issue is imprecision of the estimation. In this paper, several existing methods for software effort estimation are illustrated and their aspects will be discussed. Comparing the features of the methods could be applied for clustering based abilities; it is also useful for selecting the special method for each of the projects. The history of using each estimation model leads to have a good choice for an especial software project. In this paper an example of software cost estimation is also presented in an actual software project

International Journal of Computer Theory and …, 2009

Software cost estimation is one of the biggest challenges in these days due to tremendous completion. You have to bid so close so that you can get the consignment if your cost estimation is too low are too high in that cases organization has to suffer that why it becomes very crucial to get consignment. One of the important issues in software project management is accurate and reliable estimation of software time, cost, and manpower, especially in the early phase of software development. Software attributes usually have properties of uncertainty and vagueness when they are measured by human judgment. A software cost estimation model incorporates fuzzy logic can overcome the uncertainty and vagueness of software attributes. However, determination of the suitable fuzzy rule sets for fuzzy inference system plays an important role in coming up with accurate and reliable software estimates. The objective of our research was to examine the application of applying fuzzy logic in software cost estimation that can perform more accurate result. In fuzzy logic there are various membership function for example Gaussian, triangular, trapezoidal and many more. Out of these by hit and trial method we find triangular membership function (MF) yields least MRE and MMRE and this MRE must be less than 25%. In our research this value came around 15% which is very fair enough to estimate. Cost can be found out using the equation if payment is known Cost = Effort * (Payment Month). Therefore the effort needed for a particular software project using fuzzy logic is estimated. In our research NASA (93) data set used to calculate fuzzy logic COCOMO II. From this table size of code and actual effort has been taken. In end after comparing the result we found that our proposed technique is far superior to base work.

Procedia Technology, 2013

Software development effort estimation (SDEE) has been the focus of research in recent years. No single software development estimation technique is best for all situations and linear regression (LR) has frequently been used for both small and industrial software projects. Fuzzy logic (FL) has been applied as an alternative technique to SDEE using a Mamdani Model. In order to compare the estimation accuracy of the Mamdani and Takagi-Sugeno fuzzy systems with that of an LR model, a sample of small projects was used to generate two FL models and an LR equation. Then the FL models and the LR equation were validated by estimating the effort of projects elaborated by other developers. This latter group of projects was subdivided into projects with Effort<100 and Effort 100 (as it has been demonstrated that the estimation accuracy depends on the effort, which is an amount of time in human-hours). The results showed that the Takagi-Sugeno fuzzy system was more accurate than the Mamdani system and the LR model for SDEE of projects with Effort 100. It can be concluded that a Takagi-Sugeno fuzzy system can be useful for estimating the effort of projects with Effort 100 when they have been individually developed on a disciplined process.