Faster MIL-based Subgoal Identification for Reinforcement Learning by Tuning Fewer Hyperparameters.
Saim SunelErkin ÇildenFaruk PolatPublished in: ACM Trans. Auton. Adapt. Syst. (2024)
Keyphrases
- hyperparameters
- reinforcement learning
- model selection
- cross validation
- bayesian framework
- random sampling
- closed form
- bayesian inference
- support vector
- parameter settings
- gaussian process
- em algorithm
- maximum likelihood
- prior information
- supervised learning
- sample size
- multiple instance learning
- regularization parameter
- gaussian processes
- machine learning
- maximum a posteriori
- noise level
- learning algorithm
- incremental learning
- state space
- training set
- learning process
- image processing
- parameter optimization
- noisy images
- missing values
- incomplete data
- image segmentation
- computer vision