基于LSTM神经网络算法互联网电视EPG业务隐患预测的应用研究OA

With the deepening of the integration of the triple play and the development of smart Internet of Things technology,home broadband and internet television have become a new entry point for smart homes.In order to solve the problem of congenital lag in the timing of internet television business quality monitoring methods that cannot detect hidden faults before users,a LSTM neural network algorithm is introduced to achieve data injection intelligence business operation and maintenance capabilities.It focuses on the quality of EPG business services,identifies anomalies from business historical fluctuations and making predictive early warnings,and achieves the identification and prediction of EPG business quality hidden dangers.The length of hidden dangers discovery time is reduced to 0.5 hours,and the timely and accurate rates of hidden danger identification are above 90%.