Spectral Analysis of Covid-19 Trend in India and Its Prediction Using Locally Stationary Wavelet Process

Wavelet transform is a new analytic tool and widely used for extracting spectral information of non-stationary and chaotic signals. In discrete wavelet transform (DWT) a signal is decomposed at each stage into its average and differential behaviour. In stationary wavelet transform (SWT) the wavelet coefficients are not decimated at each stage and the translation invariance of DWT is modified. Wavelet prediction is obtained using locally stationary wavelet process. Trend is the slowest part of any signal and corresponds to the greatest scale value. Daily COVID-19 cases in India from the very beginning, March 02, 2020 to May 31, 2021 are taken as raw data. By locally wavelet prediction, it is extended up to July 25, 2021. The statistical interpretation of original and extended signal is strongly consistent with the spectral wavelet analytical results.