Heterogeneous Parallel Ensemble Learning

Abstract

Machine learning is a continuously developing field that benefits humans in enormous areas, such as systems automation, security, and medical examinations. Machine learning generally aims to extract knowledge from large masses of data and fit that data into models that can be understood and utilized. In other terms, this technology provides systems that can learn and enhance from experience automatically without being specifically programmed. In many cases, one model is not enough since models can suffer from overfitting or underfitting. Ensemble learning methods solve this issue by generating multiple models and combining the results which maintain a better prediction and lead to better performance . The goal of this research is to study, conceive and then implement a system that based on Heterogeneous ensemble learning would allow us to bypass those limits.