Final published version
Research output: Contribution to journal › Journal article › peer-review
DConfusion : A technique to allow cross study performance evaluation of fault prediction studies. / Bowes, D.; Hall, T.; Gray, D.
In: Automated Software Engineering, Vol. 21, No. 2, 04.2014, p. 287-313.Research output: Contribution to journal › Journal article › peer-review
}
TY - JOUR
T1 - DConfusion
T2 - A technique to allow cross study performance evaluation of fault prediction studies
AU - Bowes, D.
AU - Hall, T.
AU - Gray, D.
PY - 2014/4
Y1 - 2014/4
N2 - There are many hundreds of fault prediction models published in the literature. The predictive performance of these models is often reported using a variety of different measures. Most performance measures are not directly comparable. This lack of comparability means that it is often difficult to evaluate the performance of one model against another. Our aim is to present an approach that allows other researchers and practitioners to transform many performance measures back into a confusion matrix. Once performance is expressed in a confusion matrix alternative preferred performance measures can then be derived. Our approach has enabled us to compare the performance of 600 models published in 42 studies. We demonstrate the application of our approach on 8 case studies, and discuss the advantages and implications of doing this.
AB - There are many hundreds of fault prediction models published in the literature. The predictive performance of these models is often reported using a variety of different measures. Most performance measures are not directly comparable. This lack of comparability means that it is often difficult to evaluate the performance of one model against another. Our aim is to present an approach that allows other researchers and practitioners to transform many performance measures back into a confusion matrix. Once performance is expressed in a confusion matrix alternative preferred performance measures can then be derived. Our approach has enabled us to compare the performance of 600 models published in 42 studies. We demonstrate the application of our approach on 8 case studies, and discuss the advantages and implications of doing this.
KW - fault
KW - confusion matrix
KW - machine learning
U2 - 10.1007/s10515-013-0129-8
DO - 10.1007/s10515-013-0129-8
M3 - Journal article
VL - 21
SP - 287
EP - 313
JO - Automated Software Engineering
JF - Automated Software Engineering
SN - 0928-8910
IS - 2
ER -