Reference [17]

[wongsihan]

Thank you for sharing your research. I have a question regarding the code implementation of Reference [17]. In your paper, it is mentioned that the method 2 follows the construction outlined in Equation 8.
elif method == 1: lamb_hat = maxr/meanb lambd = (1-alpha_ann)*lambd + alpha_ann*lamb_hat
The original text states:"The weight for the next epoch i s updated as the moving average of the current weight (λˆ k(τ)) and the previous weight ." lamb_hat represents the current weight, and lambd represents the previous weight. However, in this code, lambd always equals 1. How does it represent the previous weight?
The code is very useful, I'm just curious why 1 is used to represent the previous weight. Looking forward to your reply.👍🏻