The Study Of Information Lifecycle Management With RAMCloud

The data is the important supporting of recent economy of both civilians and countries and is the significant non-physical ensure of current social development.Since 2012,the whole world has generally archived to the era of big data,bringing enormous needs of which has vast volume of all features.Because of it,data storage begins to become hot spot between researchers and scholars of the whole world.This paper present a data life cycle management intelligent algorithm based on monte carlo tree search algorithm of reinforcement learning and convolutional neural network of deep learning from Google deep reinforcement learning architecture which combine monte carlo tree search algorithm and convolutional neural network with linear weighting data quality by data quality assessment algorithm datarank for memory storage.The experiment compare with one model algorithm provided by University of California San Diego and two artificial abstract minding algorithms provided by sampling from staff of industry for one hundred experiment times.It provided a independent simulation environment for keeping data dynamically product.This experiment prove that the core algorithm of this paper better than model algorithm and have 61% at least better than artificial algorithm in three targets including total importance of database tables,total stored data size,and using ratio for memory storage space.