Web Effort Estimation

Mendes, Emilia; Mosley, Nile; Counsell, Steve

doi:10.1007/3-540-28218-1_2

Cited by 11 publications

(15 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [11] the following sets of measures were compared: the two sets used in [9], and the Tukutuku measures [23]. The empirical results showed that all the measures provided good estimations in terms of MMRE, MdMRE, and Pred (25) and the study largely confirmed the results of previous works. It is worth noting that the study presented in [27] [24].…”

Section: Related Worksupporting

confidence: 79%

“…This result is confirmed by the boxplots in Figure 2, which show that C3's distribution, with the Linear kernel, is less skewed than those of the other boxplots; in addition its median is closer to zero, and its outliers are closer to the tails, when compared to the outliers for the other boxplots. As for the second validation set, MMRE, MdMRE, Pred (25), MEMRE, and MdEMRE values obtained with C6 using the Gaussian kernel were better than those obtained with the other settings. Again, this result is confirmed by the boxplots in Figure 3, which show less skewed distribution for C6 with the Gaussian kernel and a median closer to zero.…”

Section: Comparison Among Different Svr Configurationscontrasting

confidence: 47%

“…We adopted the following acronyms: Readers are referred to [20] for details on how these models were created. The summary measures MMRE, MdMRE, Pred (25), MEMRE, and MdEMRE suggest that, for both validation sets, SVR provided better results than the other techniques, even though they do not fit the thresholds suggested by Conte et al [8]. Moreover, even if all the techniques outperformed MeanEffort, only SVR and MSWR are characterized by better prediction accuracy (in terms of summary measures) than MedianEffort.…”

Section: Comparison With Other Techniquesmentioning

confidence: 92%

“…Table 6 and Table 7 report the MMRE, MdMRE, Pred (25), MEMRE, and MdEMRE obtained in [20]. We adopted the following acronyms: Readers are referred to [20] for details on how these models were created.…”

Section: Comparison With Other Techniquesmentioning

confidence: 99%

“…Oliveira [31], who was the first to apply SVR to this field, used data from 18 applications from the well-known NASA software project dataset [2]. By using a leave-one-out cross-validation, the author reported that SVR significantly outperformed both linear regression and Radial Basis Function Networks (RBFNs), in terms of the indicators MMRE and Pred (25). In the second study, the authors proposed a method based on machine learning that provided an In particular, they employed robust confidence intervals, which are not associated with the probability distributions of the errors in the training set [4].…”

Section: Related Workmentioning

confidence: 99%

See 4 more Smart Citations

Applying support vector regression for web effort estimation using a cross-company dataset

Corazza

Martino

Ferrucci

et al. 2009

2009 3rd International Symposium on Empirical Software Engineering and Measurement

View full text Add to dashboard Cite

Support Vector Regression (SVR) is a new generation of Machine Learning algorithms, suitable for predictive data modeling problems. The objective of this paper is to investigate the effectiveness of SVR for Web effort estimation, in particular when dealing with a cross-company dataset. To gain a deeper insight on the method, we carried out an empirical study using four kernels for SVR, namely linear, polynomial, Gaussian, and sigmoid. Moreover, we used two variables’ preprocessing strategies (normalization and logarithmic), and two different dependent variables (effort and inverse effort). As a result, SVR was applied using six different configurations for each kernel. As for the dataset, we employed the Tukutuku database, which is widely adopted in Web effort estimation studies. A hold-out approach was adopted to evaluate the prediction accuracy for all the configurations, using two training sets, each containing data on 130 projects randomly selected, and two test sets, each containing the remaining 65 projects. As benchmark, SVR-based predictions were also compared to predictions obtained using Manual StepWise Regression, Case-Based Reasoning, and Bayesian Networks. Our results suggest that SVR performed well, since on the first hold-out, the linear kernel with a logarithmic transformation of variables provided significantly superior prediction accuracy than all the other techniques, while for the second hold-out, the Gaussian kernel achieved significantly superior predictions than all other techniques, except for Manual StepWise Regression

show abstract