• Abhinav Bangia

Updated: Predicting India Inc Bail out from COVID-19

Updated: Apr 3

In our last post, we tried to use the power function to predict the stage of the pandemic by normalizing the data set and understanding infection behavior in different stages.

The novel corona-virus or COVID-19 is no less than a bio engineered virus to change the history of the world. Over last few days, our focus has moved to our country, India. Where as by few weeks, we have been constantly modelling the COVID-19 hit countries on their data and were really frightened looking at the multiples by which this contagion manifests human bodies.

Understanding the Multiple : In our last post, if you missed it. We mentioned how COVID infections grow proportionally to the number of people already infected and their current status i.e free, quarantined, missing. While this isn't enough, the infection grows in "raise to the power" to the number of infections. Calculating this multiple is next to impossible.

Example : If we had 10 cases today, which had a status : free, could possibly infect up to 40000 people in a day.

I am personally happy with India's data right now (if it is correct), also I am writing this from garage site which I have made my isolation place since Day 1 of case rise : 17:03:2020. We are also happy to see our nations confidence and perseverance to fight this evil which is upon us now.

Below you would find our predicted model by which we are calculating number of infections v/s the dying age of the virus w.r.t day

Containment Response Dated 23rd March 2020

Containment Response Dated 25th March 2020

Containment Response Dated 03rd April 2020

Understanding the graph above :

Day 1 : 21st March 2020

Day 47 : 7th May 2020

India-Predicted Curve (Orange Color) : This model is functioned out from Chinese model, on how the virus decayed over days of quarantine and curfew (and obviously cleaning and sanitizing every place literally). The prediction curve doesn't depend upon number of cases, but depends upon log(cases today/cases yesterday), which can be applied universally. Again if you din't understand power function. You can refer to our post on 17:03:2020 for reference.

India-Actual Curve (Blue Color) : We are daily updating the numbers we source from various media citations and take a reference number which is around ballpark to the number. Apart from slight deviation from Chinese containment model, we notice we can further strict down the lock downs to meet the Chinese containment level.

Spike on day 14 : Interesting the predicted model graph had a spike on the 14th day, which luckily matches the window period of the virus infection, making us confident in the model we built.

Relief Date for India : From the prediction above, we can observe the flattening on 40-42nd day from , exactly thrice window period of the virus to show symptoms of infection in human beings

Predicted Cases : In range of 10-15 Lakhs by end of May

Remember :

Citizens : Home stay, Good Food and Hand Wash

Government & Healthcare professionals : Trace, Identify, Isolate, Treat

We will overcome this together.

For any questions, feel free : abhinav@comolho.com


Com Olho is a leading cloud-based machine learning as a service platform. We make it easy for CXO's to understand their consumers using behaviour analytics, machine learning and data science. 

e : contact@comolho.com


  • Facebook
  • Grey LinkedIn Icon

Made with      in India