Parsimonious Neural Network and Heuristic Search Method for Software Effort Estimation Model

축약형 신경망과 휴리스틱 검색에 의한 소프트웨어 공수 예측모델

  • 전응섭 (인덕대학 소프트웨어개발과)
  • Published : 2001.04.01

Abstract

A number of attempts to develop methods for measuring software effort have been focused on the area of software engineering and many models have also been suggested to estimate the effort of software projects. Almost all current models use algorithmic or statistical mechanisms, but the existing algorithmic effort estimation models have failed to produce accurate estimates. Furthermore, they are unable to reflect the rapidly changing technical environment of software development such as module reuse, 4GL, CASE tool, etc. In addition, these models do not consider the paradigm shift of software engineering and information systems(i.e., Object Oriented system, Client-Server architecture, Internet/Intranet based system etc.). Thus, a new approach to software effort estimation is needed. After reviewing and analyzing the problems of the current estimation models, we have developed a model and a system architecture that will improve estimation performance. In this paper, we have adopted a neural network model to overcome some drawbacks and to increase estimation performance. We will also address the efficient system architecture and estimation procedure by a similar case-based approach and finally suggest the heuristic search method to find the best estimate of target project through empirical experiments. According to our experiment with the optimally parsimonious neural network model the mean error rate was significantly reduced to 14.3%.

소프트웨어공수 예 에 관한 전공적인 모델링의 한계점을 극복하기 위해 사례기반과 신경망 그리고 퍼지이론 및 전문가 시스템 등 인공지능 기법을 이용한 연구들이 제시되고 있다. 특히 신경망을 이용한 공수예측 모델들이 예측력에 있어서 전통적인 모델들 보다 우수한 예측결과를 제시하고 있다. 그러나 이들 신경망 모델에 있어서도 고려되어야 할 점은 입력 데이터의 노이즈와 모델 설계 및 사용에 있어서 유연성 및 요율성 측면이 제기되고 있다. 본 연구에서는 이러한 기존의 신경망모델의 효율성 향상을 위한 새로운 방안으로 최적의 축약형 모델구조와 이에 관련된 최적 사례들을 사용하기 위한 사례기반 휴리스틱 검색기법을 제시한다. 30여개의 실제로 수행된 프로젝트의예측결과를 통해 최적사례 기반 축약형 신경망 모델의 결과가 저통적인 COCOMO 모델 그리고 기존의 신경망 모델과 비교해서 예측력과 모델의 유연성이 좋아졌음은 알 수 있었다. 따라서 본 연구에서 새롭게 제시한 축약형 모델과 최적사례기반 접근 방법은 급변하는 정보시스템 패러다임하에서도 유용하게 사용될 수있을 것이다.있을 것이다.

Keywords

References

  1. Abdel-Hamid, T. K, 'Adapting, Correcting and Perfecting Software Estimates: A Maintenance Metaphor,' IEEESoftware, Vol.5, No.4, pp.15-22. Jul 1988 https://doi.org/10.1109/52.17798
  2. Barletta, B., 'An Introduction to Case-Based Reasoning,' AI Expert, pp.43-49, Mar. 1992
  3. Bergeron, F., and St-Arnaud, J. Y., 'Estimation of Information Systems Development Efforts,' Information and Management Vol.22, pp.239-254, 1992 https://doi.org/10.1016/0378-7206(92)90026-C
  4. Boehm, B. W., 'Software Engineering Economics,' Tutorial : Software Management (3rd ed.), IEEE Computer Society, pp.135-152, 1984
  5. Boehm, B. W., 'Understanding and Controlling Software Costs,' IEEE Trans. on Soft. Eng., Vol.14, No.10, pp. 1462-1477, Oct. 1988 https://doi.org/10.1109/32.6191
  6. Hill, T., O'Connor, M. L. and Remus, M. 'Artificial Neural Network Models for Forecasting and Decision Making,' International Journal of Forecasting, pp.5-15, Sep. 1994
  7. Jun. E. S, 'Using Artificial Neural Network for Software Development Efforts Estimation,' KISS Vol.3, No.1, pp. 211-224, Jan. 1996
  8. Kermerer, C. F., 'An Empirical Validation of Software Cost Estimation Models,' Commun. Of the ACM, Vol.30, No.5, pp.416-429, May. 1987 https://doi.org/10.1145/22899.22906
  9. Kermerer, C. F., 'An Empirical Validation of Software Cost Estimation Models,' Commun. Of the ACM, Vol. 30, No.5, pp.416-429, May. 1987 https://doi.org/10.1145/22899.22906
  10. Lee, Heeseok, 'A Structured Methodology for Software Development Effort Prediction Using the Analytic Hierarchy Process,' Journal of Systems Software, Vol.21, pp.179-186, HID https://doi.org/10.1016/0164-1212(93)90040-5
  11. Lee, Jae Kyu and Kim, Ho Dong, 'Man-hours Requirement Estimation for Assemblies using Neural Networks,' 94 Japan/Korea Joint Conference on Expert Systems, Tokyo, Mar. pp.22-24, 1994
  12. Main, J., Dillon, T. S. and Khosla, R., 'Use of Neural Network for Case-Retrieval in a System for Fashion Shoe Design,' Industrial and Engineering Applications of AI and Expert Systems, Proceedings of the 8th International Conference, pp.151-158, 1995
  13. Mukhopadhyay, T., Vicinanza, S. S., and Prietula, M. J., 'Examming the Feasibility of a Case- Based Reasoning Model for Software Effort Estimation,' MISQ (4 : 1), pp.155-171, Jun. 1992
  14. Saiedian, H., Zand, M., and Barney, D., 'The Strengths and Limitations of Algorithmic Approaches to Estimating and Managing Software Costs,' International Business Schools Computing Quarterly, Spring, pp.21-22, 1992
  15. Srinivasan. K. and Fisher, D., 'Machine Learning Approaches to Estimation Software Development Effort,' IEEE Trans. On Soft. Eng., Vol.21, No.2, pp,126-137, Feb. 1995 https://doi.org/10.1109/32.345828
  16. Venkatachalam, A. R., 'Software Cost Estimation Using Artificial Neural Networks,' in Proceedings of 1993 International Joint Conference on Neural Networks, pp. 987-990, Jul. 1993 https://doi.org/10.1109/IJCNN.1993.714077
  17. Vicinanza, S. S., Mukhopadhyay, T., and Prietula, M. J., 'Software- Effort Estimation: An Explolatory Study of Expert Performance,' Information Systems Research, Vol.2, No.4, pp.243-262, Dec. 1991
  18. Wrigley, C. D., and Dexter, A. S., 'A Model for Measuring Information System Size,' MISQ, Vol.15, No.2, pp.245-257, Jun,1991 https://doi.org/10.2307/249386