Searcharxiv⌕ Search

arXiv subjects

Hussein Abbas

Publications and source records attributed to Hussein Abbas.

2 recordsLinked to original sources

Independent set reconfiguration is hard at treewidth four

Independent Set Reconfiguration asks whether one of two given independent sets of equal size can be transformed into the other by moving one token at a time while keeping the occupied vertices independent. In token sliding, a token moves along an edge. In token jumping, it may move to any unoccupied vertex. For both rules, the problem is solvable in polynomial time on trees and is PSPACE-complete on graphs of treewidth at most $c$ for some fixed constant $c$. For each rule, the smallest treewidth bound that suffices for hardness has long been open. We narrow this gap by proving that both versions are NP-hard on graphs of treewidth four.

cs.CC↗

Scalable Adversarial Online Continual Learning

Adversarial continual learning is effective for continual learning problems because of the presence of feature alignment process generating task-invariant features having low susceptibility to the catastrophic forgetting problem. Nevertheless, the ACL method imposes considerable complexities because it relies on task-specific networks and discriminators. It also goes through an iterative training process which does not fit for online (one-epoch) continual learning problems. This paper proposes a scalable adversarial continual learning (SCALE) method putting forward a parameter generator transforming common features into task-specific features and a single discriminator in the adversarial game to induce common features. The training process is carried out in meta-learning fashions using a new combination of three loss functions. SCALE outperforms prominent baselines with noticeable margins in both accuracy and execution time.

cs.LG↗