Hi..we would like to implement Rayliegh model for defect prediction in some of our projects. Just to brief on Rayleigh technique: It gives the distribution of defects across phases of the project as function of time. The key input for the function would be total estimated number of defects.
For this defect estimation we have a number of methods and one sutiable method we found to be regression analysis based on previous projects history. We found some tools in market which use regression analysis for defect estiamtion & Prediction like QSM's SLIM.
If anyone has used Rayliegh model, please share your experience. Please suggest how you have come with the defect estimation. Also suggest if you have come across some good books on defect estimation using regression anlysis. Saving Changes...
Anonymous
I practiced defect prediction while working for Lockheed Martin in the Aerospace industry. Probably the most important lessons learned were normalizing the data (different projects have different types of defects) and not over complicating the process or the analysis. In the end you may be attempting to predict the number using fairly complex regression analysis tools, but the prediction will be worthless if your data is bad or the process in getting the defect data is bad. Saving Changes...
"This planet has - or rather had - a problem, which was this: most of the people living on it were unhappy for pretty much of the time. Many solutions were suggested for this problem, but most of these were largely concerned with the movements of small green pieces of paper, which is odd because on the whole it wasn't the small green pieces of paper that were unhappy."
