Impact of COVID-19 on Forecasting Stock Prices: An Integration of Stationary Wavelet Transform and Bidirectional Long Short-Term Memory

<div>Folded RNN architecture and the process of unfolding with T time steps (x: input vector; Wxh: weight matrix between input and hidden layers; h: hidden state; Whh: weight matrix between two hidden states; Why: weight matrix between hidden and output layers; y: output vector).</div>

Complexity

fig3

Figure 3

Figure 3: Impact of COVID-19 on Forecasting Stock Prices: An Integration of Stationary Wavelet Transform and Bidirectional Long Short-Term Memory 