Updates as of July 2021 (Dynamic Boolean Network)
Published in
Jul 22, 2021
- Since the previous work is only for expository purposes, the static network, which causes termination, is now dynamic, consisting of 2 steps per iteration:
Recall that once a “backward hierarchical fixing” fails, the following two methods are added:
a. Simplification, where neurons are removed or pruned while maintaining exact equivalence of the unmodified network.
b. Augmentation, where the network adds neurons and still maintaining equivalence of the unmodified network. - Allowing architectural search poses challenge to achieving high accuracy and also constrained by using lower Boolean variables per node.
- A simple training scheme is developed by preserving last model and augmenting new neurons into it.