Nonlinear Cosine Neighborhood Time Series-Based Deep Learning for the Prediction and Analysis of COVID-19 in India

Algorithm 1

Nonlinear regressive feature selection.

Input: number of cases ‘,’ number of deaths ‘,’ and number of recovered ‘’

Output: relevant features ‘’

1: Initialize data representations ‘’, model function ‘’

2:   Begin

3:      For each case, ‘,’ deaths ‘,’ and recovered ‘’

4:         Evaluate active case ‘’ using equation (1)

5:         Evaluate the sum of squares ‘’ using equation (2)

6:         Evaluate partial derivatives for ‘’ data representations using equation (4)

7:         Obtain relevant features using equation (6)

8:         Return relevant features ()

9:       End for

10:   End