Reinforcement Learning Based Handwritten Digit Recognition with Two-State Q-Learning.
Abdul Mueed HafizGhulam Mohiuddin BhatPublished in: CoRR (2020)
Keyphrases
- reinforcement learning
- state space
- handwritten digit recognition
- function approximation
- reinforcement learning algorithms
- state action
- model free
- optimal policy
- learning algorithm
- multi agent reinforcement learning
- markov decision processes
- action selection
- stochastic approximation
- hidden state
- state abstraction
- transition model
- partially observable
- state variables
- multi agent
- action space
- temporal difference learning
- learning rate
- policy iteration
- learning agent
- neural network
- reinforcement learning methods
- temporal difference
- continuous state spaces
- markov chain
- cooperative
- hierarchical reinforcement learning
- td learning
- eligibility traces