Robot Training and Navigation through the Deep Q-Learning Algorithm.
Madson Rodrigues LemosAnne Vitoria Rodrigues de SouzaRenato Souza de LiraCarlos Alberto Oliveira de FreitasVandermi João da SilvaVicente Ferreira de LucenaPublished in: ICCE (2021)
Keyphrases
- learning algorithm
- training examples
- obstacle avoidance
- mobile robot
- robot navigation
- training samples
- training algorithm
- indoor environments
- supervised learning
- path planning
- training data
- active learning
- rough terrain
- potential field
- service robots
- collision free
- training process
- human robot interaction
- autonomous robots
- outdoor environments
- training set
- deep architectures
- autonomous vehicles
- training procedure
- dynamic environments
- navigation tasks
- contrastive divergence
- multilayer neural networks
- machine learning
- multi robot
- robotic systems
- humanoid robot
- mobile robotics
- learning problems
- reinforcement learning
- restricted boltzmann machine
- unlabeled data
- unknown environments
- back propagation
- data sets
- e learning
- motor learning
- visual navigation
- learning process
- online learning
- learning scheme
- learning tasks
- robot soccer
- autonomous navigation
- learning machines
- learning rate