Skip to content

We are looking to predict, visualize and act to contain the 2019 n-Corona Virus, through the power of data science.

License

Notifications You must be signed in to change notification settings

M-68/corona_prophet

Repository files navigation

Corona Prophet Prediction

GitHub stars GitHub forks made-with-python GitHub license

Binder

[UPDATES]

2019-nCoV Global Cases by Johns Hopkins

https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40299423467b48e9ecf6 Coronavirus 2019-nCoV Global Cases by Johns Hopkins CSSE

Can we build an original, comprehensive solution to help handle the crisis?

We are looking to predict, visualize and act to contain the 2019 n-Coron Virus, through the power of data science.

With the objective of understanding and minimizing epidemic spread, we devloped an method to accuractely predict and visualize the necessary features relating to n-Corona virus using Time Series Analysis and EDA.

The following dataset has been taken from Novel Corona Virus 2019 Dataset: along with custom feature engineering by extracting data from web.

Kernels

  • Mathematical Simulation of nCOV Transmission Model.
  • Feature engineeing and analysis of their effects in predicting extent of the nCOV Virus.
  • Prediction model for expected new cases in the Mainland China region.
  • Prediction model for expected new cases in the Mainland China region.

Mainland China* Death Trends

Following trend indicates daily,weekly,monthly trend in confirmed death rates in China .It was obtained using Prophet an open source tool for Time Series analysis.

*Mainland China includes SAR provinces and Hong Kong.

Uncertainty in Prediction

Prediction with uncertainty and changepoint

Long-term forecast and trends

Correlation Matrix

Features include StockPrice,Lowest/Highest Daily Temperature, Humidity(%), Currency, related search terms such as cold,etc.

Feature Relationships

We explore the relationship of various features with the spread of the disease by plotting graphs.

  • Stock Price Relationship

Stock

  • Humidity Relationship

Stock

  • Flights Search Relation

Stock

  • Lowest Temperature Relation

Stock

Simulations

Data Sources:

References and citations

[1] https://www.kaggle.com/sudalairajkumar/novel-corona-virus-2019-dataset

[2] https://www.cdc.gov/coronavirus/2019-ncov/index.html

[3]https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40299423467b48e9ecf6

[4] https://www.who.int/emergencies/diseases/novel-coronavirus-2019/technical-guidance

[5]https://www.maa.org/press/periodicals/loci/joma/the-sir-model-for-spread-of-disease-the-differential-equation-model

[6]https://facebook.github.io/prophet/

[7]https://ai.googleblog.com/2017/07/facets-open-source-visualization-tool.html

[8]https://scikit-learn.org/stable/

[9]https://seaborn.pydata.org/

[10]https://colab.research.google.com/

[11]https://github.com/CSSEGISandData/COVID-19/

About

We are looking to predict, visualize and act to contain the 2019 n-Corona Virus, through the power of data science.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published