A New Optimization Method for Power Resource Allocation Based on Data Analysis

To improve the overall utilization rate of resources and carry out unified security management of resources, a method for predicting the future task volume and resource demand of users is proposed based on the long short-term memory network model. Firstly, the training model was established according to the initial sample set of sample data information. The prediction model in the long short-term memory network model is used to predict the amount of future tasks [1]. Resource requirements are then mapped based on the calculation of future task volumes. Based on the long short-term memory network model, the method comprehensively considers the user’s historical law, time, business characteristics and other latitudes, analyzes the mapping relationship between the user’s task volume and the required resources based on the correlation analysis technology, and establishes a mapping model from the task volume to the resource demand, to realize the accurate prediction of the user’s future task volume and resource demand.