In particular, for the solution of the CHFM subproblem, we first propose a greedy method-based algorithm, and then, to tackle the issue of high computational complexity, we present a low complexity ...

We introduce a importance-driven weight allocation (IDWA) algorithm during the training process of the neural network. This algorithm constrains the values of less important weights to suppress the ...