Novel method for using Q-learning in small microcontrollers

Authors:
Samantha Jacobs;Fernando Rios-Gutierrez
Affiliations:
Georgia Southern University, Statesboro, GA;Georgia Southern University, Statesboro, GA
Venue:
Proceedings of the 51st ACM Southeast Conference
Year:
2013

Citing 8
Cited 0

A drive-reinforcement model of single neuron function: An alternative to the Hebbian neuronal model

AIP Conference Proceedings 151 on Neural Networks for Computing
Differential Hebbian learning

AIP Conference Proceedings 151 on Neural Networks for Computing
Neural network control for a closed-loop System using Feedback-error-learning

Neural Networks
Elevator Group Control Using Multiple Reinforcement Learning Agents

Machine Learning
Actor-critic models of the basal ganglia: new anatomical and computational perspectives

Neural Networks - Computational models of neuromodulation
Isotropic sequence order learning

Neural Computation
Learning with “Relevance”: Using a Third Factor to Stabilize Hebbian Learning

Neural Computation
Reinforcement learning: a survey

Journal of Artificial Intelligence Research

Quantified Score

Hi-index	0.00

Visualization

Abstract

A novel method for the utilizations of the Q-learning algorithm for small robot navigation is presented in this paper. The Q-learning algorithm usually requires a large amount of memory for its implementation. In this paper a novel method for implementing the Q-learning algorithm is proposed in order to reduce the amount of memory required. The benefit of using less memory is so that the algorithm can be implemented in systems with reduced amount of memory available, such as small mobile robots or robots that need a large size of memory devoted to other applications besides navigation. The Q-learning algorithm generates a map of the surrounding area by using inputs to sensors attached to the robot. The Q-learning algorithm developed makes a map of a static surrounding, the algorithm was developed using a BASIC Stamp.