This project is about forecasting COID 19 cases & deaths for 180 countries. This competition is hosted on kaggle on weekly basis. Here an attempt has been made to develop prediction model using the Week 4 COVID 19 dataset.
Week 4 COVID 19 Dataset - https://www.kaggle.com/c/covid19-global-forecasting-week-4/overview
There are three folders - Code, Datasets, Predictions.
The Code folder consists of two notebook files - Notebook-1-Timeseries-Groupings.ipynb, Notebook-2-Modelling-Bidirectional-LSTM.ipynb.
The Datasets folder consists of two folders- Provided & Modified. The kaggle datasets - train, test, submission files are available in this folder. The Train.csv file is processed by the Notebook-1-Timeseries-Groupings.ipynb file and a modified file - Train_Data_Groups.csv is produced which is available in Modified folder.
The Predictions Folder consists of two csv files Coronavirus_Forecast_LSTM_Model_Week_4.csvand submission.csv. The submission.csv file is in the format as requested by kaggle. These two csv files are produced by the Notebook-2-Modelling-Bidirectional-LSTM.ipynb file