Training Neural Networks to Play Backgammon Variants Using Reinforcement Learning.
Nikolaos PapahristouIoannis RefanidisPublished in: EvoApplications (1) (2011)
Keyphrases
- neural network
- reinforcement learning
- training process
- temporal difference learning
- training algorithm
- backpropagation algorithm
- game playing
- feed forward neural networks
- feedforward neural networks
- multi layer perceptron
- function approximation
- supervised learning
- training patterns
- test set
- training phase
- neural network model
- multilayer perceptron
- state space
- training set
- error back propagation
- neural network training
- genetic algorithm
- fuzzy logic
- neural nets
- neural network structure
- self organizing maps
- training data
- training examples
- learning capabilities
- function approximators
- learning algorithm
- learning rate
- e learning
- pattern recognition
- human players
- hidden markov models
- sufficient conditions
- learning classifier systems
- markov decision processes
- radial basis function
- hidden layer