Software Effort Estimation Using Fuzzy Logic: A Review

Software Effort Estimation Using Fuzzy Logic: A Review

Rahul Kumar Yadav Mewar University Gangrar, Chhittorghara, India

Dr. S. Niranjan

PDM School of Technology & Management, Bahadurgarh, India.

Abstract

One of the major problems with software project management is the difficulty to predict accurately the required effort for developing software applications. This is due to the reason that most of the software estimates should be performed at the beginning of the life cycle, when we do not yet know the problem we are going to solve. The task of effort estimation is challenging and is an important area of research in the field of Software Project Management. Though a number of estimation models exist for effort prediction, still many newer models are being proposed and active research is going on to obtain more accurate estimation models. In this paper we survey the most common and widely used effort estimation techniques using fuzzy logic. The survey shows that fuzzy logic effort estimation can be coupled with other techniques such as neural network, Bayesian Network and Particle Swarm Optimization technique. Recent trends on effort estimation have also been discussed at length.

Keywords- Software Development Effort, Effort Estimation, Fuzzy Logic Techniques, Estimation Models.

1. Introduction

   It is ideally desirable that the improvement in estimation techniques currently available to project managers would facilitate increased control of time and overall cost benefit in software development life cycle. Furthermore, any improvement in the accuracy of predicting the development effort can significantly reduce the costs from errors, such as estimating inaccurately, inappropriate tendering bids, and disabling the monitoring progress. Software

   development effort estimates are the basis for project bidding and planning. The consequences of poor budgeting and planning can be disastrous: if they are too pessimistic, business opportunities can be gone astray, while optimism may be followed by significant loss. Software effort estimation has even been identified as one of the three most demanding challenges in software application areas [1]. During the development process, the cost and time estimates are useful for the initial rough validation and monitoring of the projects completion process. And in addition, these estimates may be useful for project productivity assessment phases. Software effort estimation models are divided into two main categories: viz., algorithmic and non-algorithmic. The most popular algorithmic estimation models include Boehms COCOMO [2], Putnams SLIM[3] and Albrechts Function Point[4]. Non-algorithmic techniques include Price-to-Win [1],Parkinson [1], expert judgment [1] and machine learning approaches[5]. Machine learning is used to group together a set of techniques that embodies some of the facets of human mind [5]. For example, fuzzy systems, analogy, regression trees, rule induction and neural networks are among the machine learning approaches, and fuzzy systems and neural networks are considered to belong to the soft computing paradigm.

   1. Algorithmic models

      Some of the famous algorithmic models are: Boehms COCOMO81, II (Boehm et al., 2000), Albrechts Function Point (Boehm et al., 2000; Boehm, 1995) and Putnams (1978) SLIM. All of them require inputs, accurate estimate of specific attributes, such as Line of Code (LOC), number of user screen, interfaces and complexity, which are not easy to acquire during

      the early stage of software development life cycle process. Models based on historical data have limitations. Understanding and the calculation using these models are difficult due to inherent complex relationships between the related attributes, which are unable to handle categorical data as well as lack of reasoning capabilities [6]. Besides, attributes and relationships used to predict software development effort those could change with the passage of time and/or differ for software development environments (Srinivasan and Fisher, 1995). The limitations of the algorithmic models led to the exploration of the nonalgorithmic techniques visualised through soft computing philosophy.

   2. Non-Algorithmic models

      In 1990s non-algorithmic model was conceptualized and have been proposed to project cost estimation. Software researchers have turned their attention to new approaches those are based on soft computing methodologies such as based on artificial neural networks and fuzzy logic models and genetic algorithms based implementations. Neural networks are able to generalize from trained data set. A set of training data, a specific learning algorithm makes a set of rules that fit the data and fits previously unseen data in a rational manner as well. Some of the early works show that neural networks are adequately applicable to cost estimation phases as presented in the works of Venkatachalam [7] and Krishna and Satsangi [8]. Fuzzy logic offers a powerful linguistic representation that is sufficiently accommodate the imprecision in inputs and outputs, while providing a more realistic knowledge based approach to model building. Contemporary research establishes to some extent that fuzzy logic model achieved good performance index, being outperformed in terms of accuracy only by neural network model with considerably more input variables. Hodgkinson and Garratt in their works presented that estimation by expert judgment was better than all regression based models [9].

   3. Fuzzy logic models

      A fuzzy model is used when the systems are not suitable for analysis by conventional approach or when the available data is uncertain, inaccurate or vague [10]. The fuzzy model uses the fuzzy logic concepts introduced by Lofti A. Zadeh [11]. Fuzzy reasoning consists of three main components [12]: fuzzification process, inference from fuzzy rules and defuzzification process. Fuzzification process is where the objective term is transformed into a fuzzy concept. The membership functions are applied to the actual values

      of variables to determine the confidence factor or membership value (MV). Fuzzification allows input and output to be expressed in linguistic terms. Inferencing involves defuzzification of the conditions of the rules and propagation of the confidence factors of the conditions to the conclusion of the rules. Defuzzification process refers to the translation of fuzzy output into objective terms.

      A system based on Fuzzy Logic has a direct relationship with fuzzy concepts (such as fuzzy sets, linguistic Variables etc.) and fuzzy logic. The popular fuzzy logic systems can be categorised into three types: pure fuzzy logic systems, Takagi and Sugenos fuzzy system, fuzzy logic systems with fuzzification and defuzzification [12]. Since most of the engineering applications produce crisp data as input and expects crisp data as output, the last type i.e., fuzzy logic system with fuzzification and defuzzification is most widely used one and was first proposed by Mamdani. It has been successfully applied to a variety of industrial processes and consumer products [12].

      1.3.1 Fuzzy Logic in Software Effort Estimation

      A fuzzy set theoretic model is a modelling construct featuring two main properties [13]: (1) It operates at a level of linguistic terms (fuzzy sets), and (2) it represents and processes uncertainty. Fuzzy logic offers a particularly convenient way to generate a keen mapping between input and output spaces thanks to the natural expression of fuzzy rules. In software development effort estimation, two considerations justify the deciion of implement–ing a fuzzy model:1) it is impossible to develop a precise mathematical model of the domain [14]; second, metrics only produce estimations of the real complexity. Thus, according to the previous assertions, formulating a tiny set of natural rules describing underlying interactions between the software metrics and the effort estimation could effortlessly reveal their intrinsic and wider correlations.

2. Review of Software Estimation Based On Fuzzy Logic Techniques

   During the last decade, many methodologies have been developed in the areas of software cost estimation for improving estimation accuracy. Here we present a tabular view (Table 1) of works of various authors on software development effort estimation based on Fuzzy Logic techniques and concepts.

   Table 1. Research on Software Development Effort Estimation Based On Fuzzy Logic Techniques.

   Authors	Year	Related Work Done	Result Reported
   Fei Z and Liu X	1992	Introduced the f-COCOMO model	Since there was no comparison of
   [15]		which applied Fuzzy Logic to the COCOMO model for software effort	results between the f-COCOMO and other effort estimation models in their
   		estimation.	study the estimation capability of their
   			model is unknown.
   S. Kumar, B.A.	1994	Had applied fuzzy logic in	The w o r k s h o w e d h o w fuzzy
   Krishna and P.S.		Putnams manpower buildup index	F A M s can be effectively applied to
   Satsangi [16]		( MBI) estimation model. MBI	the domain of software project
   		selection process was based	management and control for the
   		upon 64 different fuzzy associative	estimation of the MBI.
   		memory (FAM) rules.
   Gray and MacDonell [17]	1997	Compared Function Point Analysis, Regression techniques, feedforward neural network and fuzzy logic in software effort estimation.	Their results showed that fuzzy logic model achieved good performance, being outperformed in terms of accuracy only by neural network model with considerably more input variables.
   Gray and	1999	Developed FULSOME (Fuzzy	The automatically generated fuzzy
   MacDonell [18]		Logic for Software Metrics) which is a	model performs acceptably when
   		set of tools that helps in creating fuzzy	compared to regression-based models.
   		model.
   J. Ryder [19]	1998	Researched on the application of fuzzy logic to COCOMO and Function Points models.	Result showed Fuzzy Logic is good at making effort estimations.
   P. Musflek, W.	2000	Worked on fuzzifying basic COCOMO	They concluded that (a) fuzzy sets help
   Pedrycz, G. Succi		model without considering the	articulate the estimates and their
   and M. Reformat		adjustment factor. In their simple f-	essence (by exploiting fuzzy numbers
   [20]		COCOMO model, the size input into the	described by asymmetric membership
   		COCOMO model is represented by a	functions) and (b) they generate a
   		fuzzy set, while a and b coefficients are	feedback as to the given uncertainty
   		crisp values. Besides the size,	(granularity) of the results.
   		augmented f- COCOMO also fuzzified
   		both the coefficients related to the
   		developm ent mode.Triangular memb-
   		ership functions are used in this
   		study.
   A.Idri, A. Abran,	2000	Proposed fuzzy intermediate	Validation results showed that the
   L. Kjiri [21]		COCOMO'81. The FLM is based upon	fuzzy intermediate COCOMO81 can
   		trapezoidal membership functions. The	tolerate imprecision in its input (cost
   		dataset is randomly generated and	drivers) and generate more gradual
   		compared with actual data of	outputs. Thus fuzzy intermediate
   		COCOMO 81. The effort multiplier for	COCOMO81 is less sensitive to the
   		each cost driver is obtained from fuzzy	changes in the inputs as compared

   		set, enabling its gradual transition from one interval to a contiguous interval such as from high to very high).	to intermediate COCOMO81.
   A. Idri, and A.	2002	Proposed an approach based on	Taking into account their results,
   Abran[22]		fuzzy logic named Fuzzy Analogy. Its	they suggested the following ranking of the
   		dataset is that of COCOMO 81.	four techniques in terms of accuracy and
   			adequacy to deal with linguistic values: 1.
   			Fuzzy Logic, 2. Fuzzy intermediate
   			COCOMO81, 3.Classical intermediate
   			COCOMO81, and 4. Classical Analogy.
   Huang, X.,	2003	Proposed a model combining fuzzy	The results of the fuzzy logic model
   Capretz. L.F.,		logic and neural networks. The	were better than those of the COCOMO
   Ren, J., Ho. [23]		dataset was obtained from the	equations. The FLM was based
   		original COCOMO (1981).	upon triangular membership functions.
   			The main benet of this model is its
   			good interpretability by using the fuzzy
   			rules.
   M.O. Saliu, M.	2004	They fuzzyfied the two different	This approach is able to deal with
   Ahmed and J.		portions of the intermediate COCOMO	uncertainty, provides transparency on
   AlGhamdi. [24]		model i.e. nominal effort estimation	prediction rationale through rules,
   		and the adjustment fac tor . They	incorporate experts knowledge in the
   		p rop os ed a fu z z y logic framework	definition of membership functions and
   		for effort prediction by integrating the	rules, as well as adaptable to new data
   		fuzzified nominal effort and th	by changing the parameters of
   		fuzzified effort multipliers of	membership functions.
   		theintermediate COCOMO model.
   Ahmed, M.A.,	2004	Presented a FLM based upon	Results showed that the FLM was
   Saliu, M.O. and		triangular membership functions.	slightly better than COCOMO equations.
   AlGhamdi, J. [25]		The dataset for validating the	In addition, they reported promising
   		FLM was (a) generated randomly	experimental summary results in
   		and (b) that of COCOMO 81 was	spite of the little background knowledge
   		used.	of the rule base and training data.
   Crespo, F.J.,	2004	Explored fuzzy regression techniques	Fuzzy regression is able to obtain
   Sicicila, M.A.,		based upon fuzzication of input	estimation models with similar predictive
   Cuadrado, J.J.		values. Project database of COCOMO-	properties than existing basic estimation
   [26]		81 are used.	models.
   M.R. Braz, S.R.	2004	Applied Fuzzy Logic for effort	Results showed that FUSP fares
   Vergilio. [27]		estimation of object-oriented software.	better than USP.
   		FUSP (Fuzzy use case size
   		points) metric allows gradual
   		classifications of use case size
   		points in the effort estimation by using
   		fuzzy numbers.
   Xu and	2004	Presented a fuzzy identication cost estimation modelling technique to deal with linguistic data, and	It was observed that the fuzzy identication model provided signicantly better cost estimations than the three

   Khoshgoftaar [28]		automatically generate fuzzy membership functions and rules. A case of study based on the COCOMO81 database compared the proposed model with all three COCOMO81 models (basic, intermediate and detailed).	COCOMO81 models.
   L.M.Cuauhtemoc, Y.M. Cornelio and G.T.Agustin. [29]	2006	Carried out a study to compare personal Fuzzy Logic Systems (FLS) with linear regression using evaluation criteria which is based upon ANOVA of MRE and MER, as well as MMRE, MMER and pred(25)	Results show that a FLS can be used as an alternative for estimating the development effort at personal level.
   Moon Ting Su,	2007	Proposed an enhanced fuzzy logic	The analysis of the results shows that
   Teck Chaw Ling,		model for the estimation of software	FLECE is able to obtain more accurate
   Keat Keong		development effort. The model Fuzzy	results in the estimation of software
   Phang, Chee		Logic Model for Software	development effort when compared
   Sun Liew and		Development Effort and Cost	to the previous fuzzy logic model.
   Peck Yen Man		Estimation (FLECE) possesses	Hence, the enhancements to FLECE
   [30]		similar capabilities as the previous	are truly useful and had given better
   		fuzzy logic model. In addition to that,	performance to the model.
   		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).
   Venus Marza,	2008	Hybrid neuro-fuzzy technique is used	The results showed that neuro-fuzzy
   Amin Seyyedi,		for development time and is validated	system is much better than two other
   and Luiz		with gathered data.	mentioned methods (fuzzy logic and
   Fernando			neural network separately).Hence, In
   Capretz[31]			order to achieve more accurate
   			estimation, several techniques maybe
   			combined.
   Parvinder S.	2008	Neuro-Fuzzy technique is used for	The performance of the Neuro-fuzzy
   Sandhu, Porush		software estimation of NASA software	based effort estimation Model and the
   Bassi, and		project data and performance of the	other existing Halstead Model, Walston-
   Amanpreet Singh		developed models are compared with the	Felix Model, Bailey-Basili Model and Doty
   Brar[32]		Halstead, Walston-Felix, Bailey-Basili	Model models is compared for effort
   		and Doty Models	dataset .The results show that the Neuro-
   			fuzzy system has the lowest MMRE and
   			RMSSE values.

   Iman Attarzadeh and Siew

   2009

   Proposed an enhanced Fuzzy Logic approach for the estimation of software development effort.

   Results s h o w e d t h a t t h e v a l u e of M M R E applying their Fuzzy Logic model was substantially lower than MMRE values

   Hock Ow [33]			as calculated by applying other Fuzzy Logic models.
   Ch. Satyananda Ready, KVSVN Raju[34]	2009	The proposed work is based on COCOMO dataset and the experimental part of the study illustrates the approach using Gaussian membership function	Result showed the proposed model gives more precise result than that of using the TMF. Thus by using GMF, the accuracy of effort estimation can be improved and the estimated effort can be very close to the actual effort.
   	2010	Proposed an approach combining the neuro-fuzzy technique and the SEER-	Results shows that that combining the neuro-fuzzy model with the SEER-SEM
   Wei Lin Du, Danny Ho, Luiz Fernando Capretz[35]		SEM effort estimation algorithm and evaluate the prediction performance of the proposed neuro-fuzzy model with SEER-SEM in software estimation practices.	effort estimation model produces unique characteristics and performance improvements. Results also proves that the proposed neuro-fuzzy structure can be used with other algorithmic models besides the
   			COCOMO model.
   Abou Bakar Nauman, Romana Aziz[36]	2011	This paper proposes a simple Bayesian Network (BN), based on classification approach. The classes of ranges of size value are distributed with help of fuzzification to distribute the probability of crisp value.	The proposed model shows two specific achievements. 1). Model shows that a smaller Bayesian network can be developed to achieve intelligent effort estimates. 2). The classifications of sizes can be managed with the help of fuzzy logic.
   Prasad Reddy P.V.G.D, Sudha K. R, Rama Sree [37]	2011	Software development effort predicted using Fuzzy Triangular Membership Function and GBell Membership Function is implemented and compared with COCOMO using NASA93 dataset.	Results shows that software effort estimation using Fuzzy method with TMF (triangular membership function) is better than Fuzzy method using GBellMF or Intermediate COCOMO. It is not possible to evolve a method, which can give 100 % VAF. By suitably adjusting the values of the parameters in FIS we can optimize the estimated effort.
   A.BalaKrishna, T.K.Rama Krishna[38]	2012	The propsed work is to employ Particle Swarm Optimization for tuning the effort parameters, fuzzy logic for reducing uncertainty in input and test its	Results shows that the proposed model reduce the uncertainty in the input sizes by using fuzzy logic and by lining the parameters of the cost model using PSO
   		suitability for software effort estimation.	with inertia weight in order to generate an
   		This methodology is then tested using	optimal result. The model was proved to be
   		NASA dataset provided by Boehm. The	efficient on the basis of VARE, MARE and
   		results are then compared with the models such as Baily-Basili, Alaa F.	VAF after comparing with the models such as Baily-Basili, Alaa F. Sheta, and Harish
   		Sheta, and Harish models.	models.

3. Conclusion

   Although many researchers contributed on cost/effort estimation, still many issues on cost/effort estimation remain unresolved. In this paper we presented a review on the Fuzzy Logic applications in Software development effort estimation models development. We also discussed the various advantages of Fuzzy Logic for developing prediction models. In order to achieve more accurate estimation, voting the estimated values of several techniques and combine their results maybe be useful. Further results can explore using four fuzzy logic membership functions Fuzzy Triangular Membership Function, GBell Membership Function, Gauss Membership Function and Trapezoidal Membership Function and their results will be compared with other estimation models and actual data set of the project. The fuzzy logic models for effort estimation can be deployed on COCOMO II environment for creating an appropriate expert system for providing required information for developing fuzzy sets and an appropriate rule base.

4. References

1. Boehm B., C. Abts and S.Chulani,200. Software Development Cost Estimation Approaches- A Survey, University of Southern California Center for Software Engineering, Technical Reports, USC-CSe-200-505.

2. B. W. Boehm, Software Engineering Economics, Englewood Cliffs, NJ, Prentice Hall, 1981.

3. L. H. Putnam, A general empirical solution to the macro software sizing and estimating problem. IEEE Transactions on Software Engineering, SE-4(4), 1978, pp 345-361.

4. Attar Software. Fuzzy Logic in Knowledge Builder, White Paper, http://www.intellicrafters.com/fuzzy.htm,2002.

5. MacDonell, S.G. and A.R. Gray, 1997 A Comparison of Modeling techniques for Software development Effort Prediction, ir. Proceedings of the 1997 International Conference on Neural Information Processing and Intelligent Information Sysems, Dunedin, New Zeeland, Springer-Verlag. Pp: 869-872.

6. Boetticher, G.D., 2001. An assessment of metric contribution in the construction of a neural network- based effort estimator.

7. Venkatachalam, A.R., 1993. Software cost estimation using artificial neural networks. Proceedings of the 1993 International Joint Conference on Neural Networks, pp: 987-990.

8. S. Kumar, B.A. Krishna, and P.S. Satsangi, "Fuzzy systems and neural networks in software engineering project management. Journ- al of Applied Intelligence, Vol. 4, 1994, pp. 31-52.

9. Hodgkinson, A. C. and P.W. Garratt, 1999. A NeuroFuzzy Cost Estimator, in Proceddings of the 3rd

   International Conference on Software Engineering and Applications SAE.,. pp: 401-406.

10. R. Babuska, Fuzzy Modeling for Control, Kluwer Academic Publishers, Dordrecht, 1999.

11. Lofti Zadeh, A., 1994. Fuzzy Logic, Neural Networks and Soft Computing, Communication of ACM 37(3): 77-84.

12. Lofti Zadeh, A., 2001., The Future of Soft Computing, In Joint 9th IFSA World Congress and 20th NAFIPS International Conference, Vancouver, Canada.

13. Pedrycz, W., Gomide, F., 1998. An Introduction to Fuzzy Sets. The MIT Press.

14. Lewis, J.P., 2001. Large limits to software estimation. ACM Software Engineering Notes 26 (4), 54 59.

15. Fei Z, Liu X, f-COCOMO Fuzzy Constructive Cost Model in Software Engineering in IEEE international conference on Fuzzy systems, 1992, pp. 331-337.

16. S. Kumar, B.A. Krishna, and P.S. Satsangi, "Fuzzy systems and neural networks in software engineering project management. Journ- al of Applied Intelligence, Vol. 4, 1994, pp. 31-52.

17. A. R. Gray, S. G. MacDonell, Applications of Fuzzy Logic to Software Metric Models for Development Effort Estimation. Fuzzy Information Processing So- ciety 1997 NAFIPS 97, Annual Meeting of the North American, 21 24 September 1997, pp. 394 399.

18. S. G. MacDonell, A. R. Gray, J. M. Calvert. FULSOME: Fuzzy Logic for Software Metric Prac- titioners and Researchers in Proceedings of the 6th International Conference on Neural In- formationProcessing ICONIP99, ANZIIS99, ANNES99 and ACNN99, Perth, Western Australia, IEEE Computer Society Press, 1999, pp.308-313.

19. J. Ryder, Fuzzy Modeling of Software Effort Predic- tion in Proceeding. of IEEE Information Technology Conference, Syracuse, NY, 1-3 Sept 1998, pp: 53-56.

20. P. Musflek, W. Pedrycz, G. Succi, M. Reformat,Software Cost Estimation with Fuzzy Models, Applied Computing Review, Vol. 8, No.2, 2000, pp. 24-29.

21. A. Idri, A. Abran, L. Kjiri. COCOMO cost model using Fuzzy Logic, 7th International Conference on Fuzzy Theory & Techniques, 27 Feb 3 March 2000, Atlantic City, New Jersey.

22. A. Idri, A. Abran,, T.M. Khoshgoftaar. Estimating software project effort by analogy based on linguistic values in.Proceedings of the Eighth IEEE Symposium on Software Metrics, 4-7 June 2002, pp. 21 30.

23. Huang, X., Capretz. L.F., Ren, J., Ho, D.A., Neuro- Fuzzy Model for Software Cost Estimation. In: Proceedings of the Third International Conference on Quality Software, 2003.

24. M.O. Saliu, M. Ahmed and J. AlGhamdi. Towards adaptive soft computing based software effort prediction Fuzzy Information, 2004. Processing NAFIPS '04. IEEE Annual Meeting of the North American Fuzzy Information Processing Society, 27- 30 June 2004, Vol.1, pp. 16-21.

25. Ahmed, M.A., Saliu, M.O., AlGhamdi, J., Adaptive

    Fuzzy Logic-Based Framework for Software Development Effort Prediction, Information and Software Technology. Elsevier, 2004.

26. Crespo, F.J., Sicicila, M.A., Cuadrado, J.J., On the useof fuzzy regression in parametric software estimation models: integrating imprecision in COCOMO cost drivers. WSEAS Transactions on Systems, 2004.

27. M.R. Braz, S.R. Vergilio, Using fuzzy theory for effort estimation of object-oriented software. 16th IEEE International Conference on Tools with Artificial Intel- ligence, ICTAI 2004, 15-17 Nov 2004, pp. 196 201.

28. Xu, Z., Khoshgoftaar, T.M., 2004. Identication of fuzzy models of software cost estimation. Elsevier Fuzzy Sets and Systems 145 (1), 141163.

29. L.M. Cuauhtemoc, Y.M. Cornelio, G.T. Agustin. Fuzzy Logic Systems for Software Development Effort Estimation Based Upon Clustering of Pro- grams Segmented by Personal Practices, Elec- tronics, Robotics and Automotive Mechanics Conference (CERMA'06), 2006, pp. 367-372.

30. Moon Ting Su, Teck Chaw Ling, Keat Keong Phang, Chee Sun Liew, Peck Yen Man, Enhanced Software Development Effort And Cost Estimation Using Fuzzy Logic Model, Malaysian Journal of Computer Science, Vol. 20(2), 2007

31. Venus Marza, Amin Seyyedi, and Luiz Fernando Capretz. Estimating Development Time of Software Projects Using a Neuro Fuzzy Approach, World Academy of Science, Engineering and Technology 46 2008.

32. Parvinder S. Sandhu, Porush Bassi, and Amanpreet Singh Brar. Software Effort Estimation Using Soft Computing Techniques, World Academy of Science, Engineering and Technology 46 2008.

33. Iman Attarzadeh and Siew Hock Ow, Software Development Effort Estimation based on a New Fuzzy Logic Model, International Journal of Computer Theory and Engineering, Vol 1. No. 4 October 2009.

34. Ch. Satyananda Reddy, KVSVN Raju. An Improved Fuzzy Approach for COCOMOs Effort Estimation using Gaussian Membership Function, JOURNAL OF SOFTWARE, VOL. 4, NO. 5, JULY 2009

35. Wei Lin Du, Danny Ho, Luiz Fernando Capretz. Improving Software Effort Estimation Using Neuro- Fuzzy Model with SEER-SEM, Global Journal of Computer Science and Technology, Vol. 10 Issue 12 (Ver. 1.0) October 2010.

36. Abou Bakar Nauman, Romana Aziz. Development of Simple Effort Estimation Model based on Fuzzy Logic using Bayesian Networks, IJCA Special Issue on Artificial Intelligence Techniques – Novel Approaches & Practical Applications AIT, 2011.

37. Prasad Reddy P.V.G.D, Sudha K. R, Rama Sree. Application of Fuzzy Logic Approach to Software Effort Estimation, International Journal of Advanced Computer Science and Applications, Vol. 2, _o. 5, 2011

38. A.BalaKrishna, T.K.Rama Krishna. Fuzzy And Swarm Intelligence For Software Effort Estimation, Advances in Information Technology and Management (AITM)

,Vol. 2, No. 1, 2012, ISSN 2167-6372.