Title | ||
---|---|---|
Integrating non-parametric models with linear components for producing software cost estimations. |
Abstract | ||
---|---|---|
•Semi-parametric models (SPMs) are introduced and applied to software cost estimation (SCE).•SPMs combine linear regression and well-known non-parametric models (e.g., ANN, SVM).•A multiple comparisons statistical methodology is used for assessment.•SPMs along with non-parametric models and linear least squares regression are ranked and clustered according to their accuracy.•Results show that the use of SPMs is beneficial especially when mixed types of relations exist between cost and various predictors. |
Year | DOI | Venue |
---|---|---|
2015 | 10.1016/j.jss.2014.09.025 | Journal of Systems and Software |
Keywords | Field | DocType |
Software cost estimation,Semi-parametric models | Data mining,Computer science,Support vector machine,Local regression,Software system,Software project management,Cost estimate,Software,Artificial intelligence,Linear least squares,Machine learning,Linear regression | Journal |
Volume | Issue | ISSN |
99 | C | 0164-1212 |
Citations | PageRank | References |
3 | 0.37 | 51 |
Authors | ||
4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Nikolaos Mittas | 1 | 3 | 0.37 |
Efi Papatheocharous | 2 | 133 | 21.97 |
Lefteris Angelis | 3 | 1296 | 82.51 |
Andreas S. Andreou | 4 | 216 | 36.65 |