Enhancing software quality estimation using ensemble-classifier based noise filtering

This paper presents a technique that improves the accuracy of classification models by enhancing the quality of training data. The idea is to eliminate instances that are likely to be noisy, and train classification models on “clean” data. Our approa