Prediction
for India using Chi-Square Optimization
Dr Himanshu
Shekhar
Introduction:
Considering the tremendous daily confirmed cases in India, the chances of
turnaround seems a remote possibility for India. It is clocking close to 18000
cases daily and now, I am finding many cases, in my vicinity and to someone I know.
It is a disturbing situation, but my inquisitive mind continues to explore more
from the data available till date for the confirmed cases. I have tried
Chi-Square optimization of error, using normal distribution curve and
prediction for future is made.
Chi-Square Minimization: The
daily confirmed cases for India is taken from 15.03.2020 to 30.06.2020 and Normal
distribution curve is plotted with chosen Mean, standard deviation and Amplitude.
The Chi-square error value for each date is calculated and then the sum of
error is minimized, using iterative scheme to optimize mean, standard deviation
and amplitude. The optimized Normal distribution curve, superimposed over daily
confirmed cases is given below.
Total Confirmed Cases: From
optimized result, cumulative confirmed cases are calculated and superimposed
over available data of total confirmed cases. It is extrapolated in future till
September 2020.
The prediction is tabulated
below.
|
Cases |
Date |
|
600000 |
02-07-2020 |
|
700000 |
07-07-2020 |
|
800000 |
12-07-2020 |
|
900000 |
16-07-2020 |
|
1000000 |
20-07-2020 |
|
1500000 |
08-08-2020 |
|
2000000 |
27-08-2020 |
Conclusion:
This is a mathematical exercise and I am finding it difficult to understand,
why these numbers for India is not reducing. Hope that we the people of India wake
up with some sense to follow the norms, suitable for restriction of confirmed
cases.
Dr Himanshu Shekhar
No comments:
Post a Comment