Time Prediction Model of WeChat Official Account User Number Based on Nonlinear Autoregressive Neural Network

E-sign is an WeChat official account which also serves as a national-level university student innovation and entrepreneurship project focusing on WeChat sign-ins. Since launched, it has attracted the attention and praise of college teachers and students. This paper attempts to use the existing data of the WeChat official account to reveal the temporal growth law of the number of users, so as to predict the future development trend of it. In this paper, from the perspective of time series, applying the neural network, we constructed a prediction model. Through continuous data training, the ideal number of hidden neurons and delay days are obtained, which yields good predictive results.