2020-Graph Active Learning

  • Course project for 2020 Fall COMP 596 Network Science
  • We studied common active learning algorithms on graphs and measured their impact on node classification performance of GCN and Simplified GCN (SGC) on several benchmark datasets
  • We proposed a new type of active learning algorithm that showed superior to or on par with competitors across different datasets
  • Project report available here
Avatar
Bruce (Zhi) Wen
Master Student in Computer Science