Abstract

A number of software effort estimations have attempted using statistical models, case based reasoning, and neural networks. The research results showed that the neural network models perform at least as well as the other approaches, so we selected the neural network model as the estimator. However, since the computing environment changes so rapidly in terms of programming languages, development tools, and methodologies, it is very difficult to maintain the performance of estimation models for the new breed of projects. Therefore, we propose a search method that finds the right level of relevant cases for the neural network model. For the selected case set, the scale of the neural network model can be reduced by eliminating the qualitative input factors with the same values. Since there exist a multitude of combinations of case sets, we need to search for the optimal reduced neural network model and corresponding case set. To find the quasi-optimal model from the hierarchy of reduced neural network models, we adopted the beam search technique and devised the case-set selection algorithm. We have shown that the resulting model significantly outperforms the original full model for the software effort estimation. This approach can be also used for building any case-selective neural network.

Author Keywords: Software effort estimation; Reduced neural network model; Case set hierarchy; Beam search; Sensitivity of beam width; Case-set selection algorithm

Article Outline

1. Introduction

2. Software effort estimation models

2.1. Statistical models

2.2. Artificial intelligence models

2.3. Case-based reasoning models

2.4. Neural network models

2.5. Roads ahead

3. Full neural network model for software effort estimation

3.1. Factors for software effort estimation

3.2. Full neural network model

4. Case-selective reduced neural network models

4.1. Reduced neural network models

4.2. Measure of similarity and case sets hierarchy

4.3. Sample size effect and pruning policy

5. Search for the quasi-optimal case-selective neural network model

5.1. Non-monotonicity and beam search

5.2. Notation

5.3. Case set selection algorithm

6. Illustration of the case-set selection algorithm

7. Performance of case-set selection algorithm

7.1. Sensitivity of stopping rule

7.2. Sensitivity of the beam width

7.3. Performance of the quasi-optimal reduced model

8. Conclusion

References