Covid-19 Forecasting

Forecasting Covid-19 cases using deep learning and statistical models for few countries.

Objectives

Our main objective is to see which one of our chosen 4 countries have handled the virus in a way that can be generalized to everyone as simple guidelines, the targeted countries are

• United States
• Germany
• Italy
• South Korea
• Malaysia

Data

the first notebook uses data from our world in data GitHub repository, it contains daily cases for all countries.

Malaysia specific notebook uses data from CSSEGISandData.

Models

for the modeling, 3 differernt models are trained and compared

• Bidrectional Long Short Term Memory (BiLSTM)
• Autoregressive Integrated Moving Average (ARIMA)
• Holt's Exponential Smoothing (HES)

The statistical models are trained in incremental fashion as defined below:

• train with original train data
• predict the next value
• append the prediction value to the training data
• repeat training and appending for n times (days in this case)

This incremental technique significantly improves the accuracy by always using all data up to previous day for predicting next value unlike predicting multiple values at the same time which is not incremental.

Notebook content

There are 2 notebooks in this project, one for the first 4 countries and another separate notebook for Malaysia data analysis, the Malaysia notebook answers questions of different types of analysis (descriptive, diagnostic, predictive and prescriptive).

• Data Exploration
• Visual and Descriptive Analysis
• Modeling
• Analysis of Effectiveness of mandated lockdown
• Conclusion