Predicting the number of new cases of COVID-19 in India using Survival Analysis and LSTM

S. Aarathi, Rithika F. Johnson, Raja Praveen K N, T R Mahesh, V Vivek

20212021 Fifth International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud) (I-SMAC)26 citationsDOI

Abstract

COVID-19 has been the cause of death for thousands of people across the globe. The goal of this paper is to forecast the new COVID-19 cases in India. The other methods used to forecast COVID-19 cases fail to give results with good accuracy when they try to predict the new cases number for a long time period or when the count of daily cases reported is large since the population of a country is large. The proposed study overcomes the challenge by firstly customizing the dataset. Second, the survival analysis has been utilized to choose appropriate factors, and third, the data will be integrated into the Long Short-Term Memory Network (LSTM). With a mean absolute percentage error of 5.79 percent, data from the 30th of January, 2020, to the 16th of June, 2021, was used to determine the new cases number of every day for the next 21 days.

Topics & Concepts

Coronavirus disease 2019 (COVID-19)Long short term memoryComputer scienceStatisticsGlobeTerm (time)PopulationArtificial neural network2019-20 coronavirus outbreakSurvival analysisMean absolute errorSevere acute respiratory syndrome coronavirus 2 (SARS-CoV-2)Time seriesArtificial intelligenceData miningRecurrent neural networkMedicineMachine learningMean squared errorMathematicsInternal medicinePathologyEnvironmental healthDiseasePhysicsOphthalmologyQuantum mechanicsOutbreakInfectious disease (medical specialty)COVID-19 diagnosis using AICOVID-19 epidemiological studiesAnomaly Detection Techniques and Applications