Improvising process scheduling using machine learning

CPU process scheduling is of utmost importance in order to reduce the idle time and improve efficiency of processes and processor. In this paper, we introduce process scheduling techniques and memory layout of processes. Two types of executions are considered - individual execution and parallel execution. Then, multiple attributes of processes are taken into consideration by the Machine Learning (ML) algorithms to predict their Turn-around-time (TaT). These are used to assign appropriate time slices in Linux scheduler. Our experiments could successfully reduce TaT by this approach.