Human Factors

The emergence and spread of infectious diseases, like COVID-19, are on the rise. Can you identify patterns between population density and COVID-19 cases and identify factors that could help predict hotspots of disease spread?

Look Before You Leap!

Summary

We've identified the human factors which contribute to COVID-19 spread. We have collected data of multiple factors which might be contributing to COVID-19 spread.We are eliminating the factors which are not responsible by Exploratory data analysis and feature engineering and using those factors predicting the hotspots

How We Addressed This Challenge

Our project has generated great results so far like -

1. Population density is a major factor in 'COVID' growth.

2. Places like bus stops, railway stations, airports are helping in 'COVID' growth.

3. The places where people chose to stay at home showed lesser 'COVID' growth.

4. Average accessibility to portable water in a place is also contributing to 'COVID'

growth.

5. Air quality and thermal anomalies are not related to 'COVID' growth.

We are also working to find -

1. Does wearing mask decreases the 'COVID' growth rate?

2. Will the concept of 'vegetarianism' lessen diseases outbreaks like COVID?

Thus we have successfully identified some factors that are contributing to 'COVID' growth.

We have done this for the districts of India as India has a large number of districts and varying meteorological factors and varying culture-making great place to get unbiased data. But we are doing it for other countries as well. The goal is two covers every part of the world.

How We Developed This Project

We collected data about the following possible factors till now -

1. Population density

2. Percentage of people staying at home

3. Percentage of people going to workplaces

4. Percentage of people going to bus stops, railway stations, airports

5. Percentage of people going to shops

6. Percentage of people going to parks

7. Air quality

8. Ease of accessibility of water

9. Thermal anomalies

We have done exploratory data analysis and have applied feature extraction techniques like calculating feature importance, univariate selection, recursive feature elimination to filter out responsible factors. We extracted four important features as per the results (density, percentage of people staying at home, the percentage of people going to transit stations and water accessibility). We then used various Machine Learning models like Logistic Regression, Random Forest, K Nearest Neighbours, Support Vector Machines and Stochastic gradient descent. We have used these models to make a prediction on both the acquired data and data obtained after filtering out the above listed four factors and we found there is no major change in the prediction results which proves the extracted features are enough to predict hotspots. We have automated the process of data acquisition and prediction so it updates automatically.

Project Demo

1) For a SlideShare presentation, click here.

2)For youtube video, click here.

Data & Resources

Fire and thermal anomalies - NASA  

COVID-19 cases - Pathfinders, NASA, Google

Mobility factors - Google

Population density - Census of India

Air quality, water availability -  Ministry of home affairs, India

scrapping other websites, newspapers etc

Tags
#air quality #coronavirus #humanfactor #lockdown #creativity #dataprediction #hotspotprediction #stayathome #endcovidby2020
Global Judging
This project was submitted for consideration during the Space Apps Global Judging process.