Automatic Optimization and Allocation of Data Using Q-Learning Technique

An automatic optimization and allocation of data is a context that attains both programmability, adaptability and operates calculating variety of CPUs, GPUs. Based on the task attribute, the medium of system is obtained, i.e. (GPU or CPU) considering the set of attributes of the system with designated task, the resultant configuration of system is predicted. During training, sample tasks are driven into network to find out particular characteristics. By using Q-learning technique comparison of graphs, the resultant outcome is obtained. By calculating the mean value among the dataset, the memory utilized by the system in future is predicted.