No-Regret Reductions for Imitation Learning and Structured Prediction
Stéphane RossGeoffrey J. GordonJ. Andrew BagnellPublished in: CoRR (2010)
Keyphrases
- imitation learning
- structured prediction
- maximum margin
- learning algorithm
- support vector
- pattern classification
- hyperplane
- markov networks
- lower bound
- support vector machine
- max margin
- binary classification
- maximum likelihood
- generalization error
- multiple kernel learning
- conditional random fields
- multi task
- training data
- belief propagation
- approximate inference
- learning problems
- learning tasks
- cross validation
- svm classifier
- reinforcement learning