Framework for Task scheduling in Cloud using Machine Learning Techniques

Task scheduling plays a vital role in the function and performance of the cloud computing system. While there exist many approaches for improving task scheduling in the cloud, it is still an open issue. In this proposed framework we try to optimize the utilization of cloud computing resources by using machine learning techniques. Task scheduling algorithms can be designed for static or dynamic scenarios. The proposed framework is for the dynamic scenario. Task scheduling can consider different parameters for scheduling purposes like Makespan, QoS, energy consumption, execution time, and load balancing. We propose to apply a machine learning technique for the incoming task requests so as to classify the best suitable algorithm for the task request rather than randomly assigning the scheduling algorithm. Supervised machine learning techniques can be used here. The outcome of the proposed work leads to the selection of the best task scheduling algorithm for the input task(request).