Predicting India’s COVID-19 Third Wave with LSTM

Complete project of predicting new COVID-19 cases in the next 90 days with LSTM

India is seeing a steep rise in COVID-19 cases again! So, I thought about using the artificial recurrent neural network (RNN) architecture Long Short-Term Memory (LSTM) to predict how the COVID-19 graph will look in next 90 days (starting from 11 January 2022).

Dataset

The dataset is downloaded from ‘COVID-19 India Datasets by DataMeet’. The data is community collected, cleaned and organized from different government websites which are freely available to all the Indians.

Github Repository: https://github.com/datameet/covid19

The dataset has a Creative Commons Attribution 4.0 International Public License. The dataset is downloaded on 10 January 2022 and contains data up to the same date.

We are using the file all_totals.JSON in the data directory.

Technology

We have used the artificial recurrent neural network (RNN) architecture Long Short-Term Memory (LSTM) for this project. The CovidPredictionLSTM.ipynb file is the Jupyter Notebook file containing all of the work.

This study/project just showcases the usage of the LSTM architecture in predicting time-series data. In this case, we used the COVID-19 data from India for our study. This model does not consider transmissibility and other factors while making the predictions.

Author Info

Samrat Dutta

Github: https://github.com/SamratDuttaOfficial

Linkedin: https://www.linkedin.com/in/SamratDuttaOfficial [Hire Me]

Wisest Friends (Machine Learning) Discord: https://discord.gg/7Bx6PGVy

Buy me a coffee: https://www.buymeacoffee.com/SamratDutta