A Reduction of Imitation Learning and Structured Prediction to No-Regret Online Learning.
Stéphane RossGeoffrey J. GordonDrew BagnellPublished in: AISTATS (2011)
Keyphrases
- online learning
- imitation learning
- maximum margin
- structured prediction
- support vector
- hyperplane
- learning algorithm
- markov networks
- support vector machine
- pattern classification
- max margin
- e learning
- multi task
- active learning
- maximum likelihood
- efficient learning
- conditional random fields
- latent variables
- multiple kernel learning
- graphical models
- multi class
- feature selection