Ensemble classification technique to detect stress in IT-professionals
Deepak Chowdary Edara, Kollimarla Anusha DEVI, D. Mounika, Venkatramaphanikumar Sistla, Venkata Krishna Kishore Kolli · 2016
In these days, Employee turnover has become a major challenge in many software industries. Most often, people suffer from stress due to heavy work pressures imposed on them and competitive spirit of the work completed in their day to day lives. A survey was conducted by software professionals who work for various companies and stress on them was investigated. For this study, the PEGASOS optimization algorithm is taken for classification. An ensemble with boosting technique is used to improve the performance of the Classifier. In this boosting, each sample is chosen according to its probability distribution which is updated to the error sample proportionally. In this model, 1000 sample data is collected from IT Professionals. Based on the data collected, this model is applied and compared with some of the existing UCI datasets such as Pima-diabetes, Heart-statlog and Sonar.