Think locally, act globally: Improving defect and effort prediction models

Bettenburg,; Nagappan,; Hassan, Fikrat M.

doi:10.1109/msr.2012.6224300

Cited by 95 publications

(79 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…No learner is best for all data sets [4] since data can change over time, making prior results outdated [39]. Hence, many researchers now explore "local learners" that eschew single global conclusions in favor of more context-dependent conclusions [1], [22], [36].…”

Section: Algorithm Tuningmentioning

confidence: 99%

See 1 more Smart Citation

Beyond data mining; towards "idea engineering"

Menzies

2013

Proceedings of the 9th International Conference on Predictive Models in Software Engineering

View full text Add to dashboard Cite

show abstract

Section: Algorithm Tuningmentioning

confidence: 99%

“…We prefer the IDEA algorithm shown above (since it runs in linear time) followed by some case-based reasoning tool such as the W tool (see Figure 3). Other teams have generated clusters like Figure 2.d using recursive regression methods [1]. Regardless of how the landscape is generated, the general principle is the same:…”

Section: Landscape Miningmentioning

confidence: 99%

Beyond data mining; towards "idea engineering"

Menzies

2013

Proceedings of the 9th International Conference on Predictive Models in Software Engineering

View full text Add to dashboard Cite

show abstract

“…[1,38,48]), as opposed to other studies which focus on direct predictions of project attributes while avoiding the human element (e.g. [34,9,31,4,28]). This complements recent works which have shown that complexity metrics, e.g.…”

Section: Introductionmentioning

confidence: 99%

On the effect of code regularity on comprehension

Jbara

Feitelson

2014

Proceedings of the 22nd International Conference on Program Comprehension

View full text Add to dashboard Cite

It is naturally easier to comprehend simple code relative to complicated code. Regrettably, there is little agreement on how to effectively measure code complexity. As a result simple generalpurpose metrics are often used, such as lines of code (LOC), McCabe's cyclomatic complexity (MCC), and Halstead's metrics. But such metrics just count syntactic features, and ignore details of the code's global structure, which may also have an effect on understandability. In particular, we suggest that code regularity-where the same structures are repeated time after time-may significantly reduce complexity, because once one figures out the basic repeated element it is easier to understand additional instances. We demonstrate this by controlled experiments where subjects perform cognitive tasks on different versions of the same basic function. The results indicate that versions with significant regularity lead to better comprehension, while taking similar time, despite being longer and having higher MCC. These results indicate that regularity is another attribute of code that should be taken into account in the context of studying the code's complexity and comprehension. Moreover, the fact that regularity may compensate for LOC and MCC demonstrates that complexity cannot be decomposed into independently addable contributions by individual attributes.

show abstract

“…More recent work has been emphasising the relatively good predictive performance achieved by ensembles of learning machines (Kultur et al 2009;Minku and Yao 2013a;Kocaguneli et al 2012) and local methods that make estimations based on completed projects similar to the project being estimated (Minku and Yao 2013a;Menzies et al 2013;Bettenburg et al 2012). For instance, Regression Trees (RTs), Bagging ensembles of MultiLayer Perceptrons (Bag + MLPs) and Bagging ensembles of RTs (Bag + RTs) have been shown to perform well across several datasets (Minku and Yao 2013a).…”

Section: For See Assuming No Chronologymentioning

confidence: 99%

Which models of the past are relevant to the present? A software effort estimation approach to exploiting useful past models

Minku

Yao

2016

Autom Softw Eng

View full text Add to dashboard Cite

Software Effort Estimation (SEE) models can be used for decision-support by software managers to determine the effort required to develop a software project. They are created based on data describing projects completed in the past. Such data could include past projects from within the company that we are interested in (WC projects) and/or from other companies (cross-company, i.e., CC projects). In particular, the use of CC data has been investigated in an attempt to overcome limitations caused by the typically small size of WC datasets. However, software companies operate in non-stationary environments, where changes may affect the typical effort required to develop software projects. Our previous work showed that both WC and CC models of the past can become more or less useful over time, i.e., they can sometimes be helpful and sometimes misleading. So, how can we know if and when a model created based on past data represents well the current projects being estimated? We propose an approach called Dynamic Cross-company Learning (DCL) to dynamically identify which WC or CC past models are most useful for making predictions to a given company at the present. DCL automatically emphasizes the predictions given by these models in order to improve predictive performance. Our experiments comparing DCL against existing WC and CC approaches show that DCL is successful in improving SEE by A preliminary version of this work appeared as Minku and Yao (2012a

show abstract

Think locally, act globally: Improving defect and effort prediction models

Cited by 95 publications

References 29 publications

Beyond data mining; towards "idea engineering"

Beyond data mining; towards "idea engineering"

On the effect of code regularity on comprehension

Which models of the past are relevant to the present? A software effort estimation approach to exploiting useful past models

Contact Info

Product

Resources

About