Double Wins: Boosting Accuracy and Efficiency of Graph Neural Networks by Reliable Knowledge Distillation.
Qiaoyu TanDaochen ZhaNinghao LiuSoo-Hyun ChoiLi LiRui ChenXia HuPublished in: ICDM (2023)
Keyphrases
- neural network
- knowledge base
- high accuracy
- computational efficiency
- pattern recognition
- computational complexity
- learning algorithm
- highly accurate
- fuzzy logic
- domain knowledge
- knowledge representation
- knowledge based systems
- self organizing maps
- knowledge sharing
- neural nets
- improving classification accuracy
- computational cost
- prior knowledge
- pairwise
- knowledge acquisition
- prediction accuracy
- feature space
- decision trees
- ensemble learning