Continuous adaptive reinforcement learning with the evolution of Self Organizing Classifiers.
Danilo Vasconcellos VargasHirotaka TakanoJunichi MurataPublished in: ICDL-EPIROB (2013)
Keyphrases
- reinforcement learning
- function approximation
- neural network
- learning classifier systems
- training data
- training set
- machine learning algorithms
- action space
- feature set
- machine learning
- learning algorithm
- multiple classifiers
- reinforcement learning algorithms
- supervised classification
- continuous state
- data streams
- multi agent
- decision trees
- data sets
- learning capabilities
- multi agent reinforcement learning
- temporal difference
- model free
- markov decision processes
- classification algorithm
- svm classifier
- optimal policy
- multi class