Abstract:The aim is to explore the development trend of COVID-19 (Corona Virus Disease 2019) and predict the infectivity of 2019-nCoV (2019 Novel Coronavirus), as well as its impact on public health. First, the existing data are analyzed through data pre-processing to extract useful feature factors. Then, the LSTM (Long-Short Term Memory) prediction model in the deep learning algorithm is used to predict the epidemic situation in Hubei Province, outside Hubei nationwide, and the whole country, respectively. Meanwhile, the impact of intervention time changes on the epidemic situation is compared. The results show that the prediction results are almost consistent with the actual values. Specifically, Hubei Province abolishes quarantine restrictions after the Spring Festival holiday, and the first COVID-19 peak is reached in late February, while the second COVID-19 peak has been reached in early March. Finally, the cumulative number of diagnoses reaches 85,000 cases, with an increase of 15,000 cases compared with the nationwide cases outside Hubei under the continuous implementation of prevention and control measures. Under the prediction of the proposed LSTM model, if the nationwide implementation of prevention and control interventions is postponed by 5 days, the epidemic will peak in early March, and the cumulative number of diagnoses will be about 200,000; and if the intervention measures are implemented five days earlier, the epidemic will peak in mid-February, with a cumulative number of diagnoses of approximately 40,000. Meanwhile, the proposed LSTM model predicts the RMSE values of the epidemic situation in Hubei Province, outside Hubei nationwide, and the whole country as 34.63, 75.42, and 50.27, respectively. Under model comparison analysis, the prediction error of the proposed LSTM model is small and has better applicability over similar algorithms. The results show that the LSTM model is effective and has high performance in infectious disease prediction, and the research results can provide scientific and effective references for subsequent related research.

SIRVD-DL: A COVID-19 Deep Learning Prediction Model Based on Time-Dependent SIRVD

Deep learning infused SIRVD model for COVID-19 prediction: XGBoost-SIRVD-LSTM approach

A Prediction Model Based on Machine Learning for Diagnosing the Early COVID-19 Patients

Forecasting the Long-Term Trend of COVID-19 Epidemic Using a Dynamic Model

Predicting the Epidemics Trend of COVID-19 Using Epidemiological-Based Generative Adversarial Networks.

VOC-DL: Deep Learning Prediction Model for COVID-19 Based on VOC Virus Variants

Cooperation or Confrontation? —— Prediction of COVID-19 Pandemic Situation Via Deep Learning

Time series predicting of COVID-19 based on deep learning

A Comparative Study on Deep Learning Models for COVID-19 Forecast

Predicting the spread of COVID-19 in Delhi, India using Deep Residual Recurrent Neural Networks

Predicting the Number of COVID-19 Cases Based on Deep Learning Methods

Temporal Deep Learning Architecture for Prediction of COVID-19 Cases in India

Prediction of the COVID-19 infectivity and the sustainable impact on public health under deep learning algorithm

COVID-19 Probability Prediction Using Machine Learning: An Infectious Approach

Coverage Prediction of Novel Coronavirus Vaccine Based on Deep Learning Time Series Analysis

TW-SIR: time-window based SIR for COVID-19 forecasts

Deep Learning for Virus-Spreading Forecasting: a Brief Survey

Predicting Infectious Disease Using Deep Learning and Big Data

A Multivariate Deep Learning Model with Coupled Human Intervention Factors for COVID-19 Forecasting.

Deep learning via LSTM models for COVID-19 infection forecasting in India