A Semantic Loss Function for Deep Learning with Symbolic Knowledge.
Jingyi XuZilu ZhangTal FriedmanYitao LiangGuy Van den BroeckPublished in: ICML (2018)
Keyphrases
- loss function
- deep learning
- symbolic knowledge
- pairwise
- artificial neural networks
- unsupervised learning
- support vector
- cognitive architecture
- machine learning
- mental models
- semantic information
- neural network
- domain specific
- weakly supervised
- natural language
- supervised learning
- state space
- knowledge representation
- object recognition
- actionable knowledge