Learning to Deceive Knowledge Graph Augmented Models via Targeted Perturbation.
Mrigank RamanSiddhant AgarwalPeifeng WangAaron ChanHansen WangSungchul KimRyan A. RossiHandong ZhaoNedim LipkaXiang RenPublished in: CoRR (2020)
Keyphrases
- prior knowledge
- learning systems
- learning algorithm
- knowledge acquisition
- learning models
- background knowledge
- reinforcement learning
- qualitative models
- knowledge transfer
- domain knowledge
- learning process
- knowledge level
- probabilistic model
- bipartite graph
- complex domains
- structured prediction
- learned models
- previously learned
- online learning
- machine learning algorithms
- concept maps
- active learning
- organizational learning
- expert systems
- fully connected