A learning algorithm based on $λ$-policy iteration and its application to the video game "tetris attack".
Hoang Thanh LePublished in: GI-Jahrestagung (2016)
Keyphrases
- video games
- policy iteration
- learning algorithm
- reinforcement learning
- markov decision processes
- model free
- reinforcement learning algorithms
- policy evaluation
- sample path
- optimal policy
- learning experience
- temporal difference
- fixed point
- computer games
- infinite horizon
- game playing
- least squares
- average reward
- game play
- finite state
- educational games
- supervised learning
- function approximation
- machine learning algorithms
- learning process
- training data
- machine learning
- optimal control
- image sequences
- linear programming
- game design
- learning tasks
- state space
- search algorithm
- learning environment
- average cost
- control system
- game players