The datasets used for analysis and prediction is taken from https://ourworldindata.org/coronavirus/country/india
The available dataset is cleaned to remove outliers which might cause deviations in the results obtained. Statistical Analysis is applied to understand the nature of the curve and to derive its properties.
The obtained curve pattern is a Mixture of Gaussians.
On application of Estimation Maximization Algorithm, the points where peaks can be expected (mean), the spread of wave (standard deviation) and the weights of each wave can be estimated.
Day 0: Indicates the beginning of Wave 1
Day 376: Indicates the beginning of Wave 2
Day 678: Indicates the beginning of Wave 3
The ML Model used analyses the trend in the 1st and the 2nd wave. The learned patterns are applied by the machine to plot the prediction made for the third wave.
The predicted wave has a shape that coincides with the actual third wave except for the slight difference in peaks.
Numerous factors to be considered that affects the crowd and the intensity of the wave are to be updated and keenly computed to avoid deviations.
The dataset is split into training and testing sections. Once a trained model is obtained, the testing data is used to compare the predicted curve with the actual curve of the second wave. More the data, better the amount of variations and information availability for the Machine to Learn and generate patterns!
Further waves can be predicted from the using the trained model given that the increase in vaccination rates and immunity against the wave are considered and explored extensively!
Why Prediction??
Awareness and Caution to the citizen's of the country and the Government ruling them!