Drop-Bottleneck: Learning Discrete Compressed Representation for Noise-Robust Exploration.
Jaekyeom KimMinjung KimDongyeon WooGunhee KimPublished in: CoRR (2021)
Keyphrases
- supervised learning
- learning systems
- learning algorithm
- learning tasks
- neural network
- learning process
- learning problems
- image noise
- reinforcement learning
- online learning
- continuous domains
- salt pepper
- data sets
- multiple representations
- visual representation
- learning analytics
- background knowledge
- higher education
- unsupervised learning
- bayesian networks