Defect prediction for new products during the development phase

Tao Peng Chu, Fu Kwun Wang*

*Corresponding author for this work

Research output: Contribution to journalJournal Article peer-review

1 Scopus citations


Defect prediction is an important task in the development phase of new products. We present a new method of predicting the number of defects for new products in the computer industry, such as laptops, during the development phase. The proposed approach is based on robust regression and reliability growth models. We use a robust regression model to predict the number of defects for new products, and a growth model is used to track the test progress during the development phase. The Weibull growth model achieves the best fit compared with other models, such as the exponential, three-parameter logistic, and Gompertz model. The predicted defects for four example products are similar to the actual defects during the whole development phase. The Weibull growth curve also provides a good fit for monitoring defect trends.

Original languageEnglish
JournalJournal of Testing and Evaluation
Issue number4
StatePublished - 07 2014
Externally publishedYes


  • Defect prediction
  • Growth model
  • Robust regression
  • Test

Cite this