Prediction Method for Communication Capability of Grassroots Intelligent Integrated Media Systems Based on Neural Networks
with the increasingly widespread application of grassroots smart integrated media systems in various regions,in order to achieve the prediction of the system's long-term communication ability,a neural network-based method for predicting the communication ability of grassroots smart integrated media systems is designed.Design efficient data min-ing algorithms based on improved dataset structure,and implement the mining of grassroots intelligent integrated media system structure,time,text,and user information.For mining information,design an anomaly event detection model to implement data anomaly event detection,and use a differential comprehensive moving average autoregressive model to re-pair single point noise.Construct a context dependent dynamic graph attention network as a grassroots intelligent inte-grated media system propagation ability prediction model,consisting of a modeling propagation dynamic graph module,a temporal spatial dependency learning module,and a prediction module.The input is processed information,and the out-put information is the prediction increment of the propagation scale of the basic intelligent integrated media system.The experimental results show that the overall accuracy of this method in predicting user lists is high,reaching a maximum of 98.95%,and the overall recall rate of predicting user lists is high,reaching a maximum of 97.24%.
Grassroots intelligent integrated media systemEnd-to-end neural networksSingle point noise restora-tionPrediction of communication ability
NETL
NSTL
万方数据
