Prediction of Epidemic Transmission Index Based on Adjusted_SEIR and LSTM Network Model

In order to make a more accurate prediction of the COVID-19 transmission index, Adjusted-SEIR model is proposed, which takes into account the classification of susceptible persons, exposed persons, infected persons and recovered persons, taking into account the influencing factors such as infection rate, contact rate and recovery rate. At the same time, this paper introduces LSTM (Long Short Term Memory networks) model, a deep learning method that can capture the dependency of time series, and analyzes and predicts the COVID-19 epidemic by combining the relevant data information of a city's epidemic prevention system. Using the historical data of the novel coronavirus epidemic as input sequence, the network model is trained to learn the dynamic pattern of epidemic transmission, and the model is used to predict the future epidemic transmission capacity .From the experimental results, compared with the traditional single machine learning algorithm model, such as Ridge regression, Lasso regression, and ordinary SEIR(Susceptible Exposed Infectious Recovered model) model, Adjusted-SEIR and LSTM models have potential advantages in predicting the spread of COVID-19,The R
²
of the Adjusted SEIR model reached 0.81.This research provides new methods and tools for outbreak prediction and decision-making, helping to improve the effectiveness of public health measures and responses to outbreak challenges.