Genetic Reinforcement Learning for Neurocontrol Problems

@article{Whitley2004GeneticRL,
  title={Genetic Reinforcement Learning for Neurocontrol Problems},
  author={L. D. Whitley and Stephen Dominic and Rajarshi Das and Charles W. Anderson},
  journal={Machine Learning},
  year={2004},
  volume={13},
  pages={259-284}
}
Empirical tests indicate that at least one class of genetic algorithms yields good performance for neural network weight optimization in terms of learning rates and scalability. The successful application of these genetic algorithms to supervised learning problems sets the stage for the use of genetic algorithms in reinforcement learning problems. On a simulated inverted-pendulum control problem, “genetic reinforcement learning” produces competitive results with AHC, another well-known… 
Training Reinforcement Neurocontrollers Using the Polytope Algorithm
A new training algorithm is presented for delayed reinforcement learning problems that does not assume the existence of a critic model and employs the polytope optimization algorithm to adjust the
Evolutionary Computation for Reinforcement Learning
TLDR
Research is surveyed on the application of evolutionary computation to reinforcement learning, overviewing methods for evolving neural-network topologies and weights, hybrid methods that also use temporal-difference methods, coevolutionary methods for multi-agent settings, generative and developmental systems, and methods for on-line evolutionary reinforcement learning.
Neuroevolution for reinforcement learning using evolution strategies
  • C. Igel
  • Computer Science
    The 2003 Congress on Evolutionary Computation, 2003. CEC '03.
  • 2003
TLDR
The CMA-ES is applied to the optimization of the weights of neural networks for solving reinforcement learning problems, and results with fixed network topologies are significantly better than those reported for the best evolutionary method so far.
Evolutionary reinforcement learning for neurofuzzy control
TLDR
This paper represents a novel approach to globally optimised and on-line learning fuzzy controllers for cases where supervised learning is diff icult and is based on the functional reasoning of fuzzy logic combined with reinforcement learning paradigm.
Reinforcement Learning with Modular Neural Networks for Control
TLDR
The reinforcement learning algorithm called Q-Learning is modified to train a modular neural network to solve a control problem and the resulting algorithm is demonstrated on the classical pole-balancing problem.
Efficient reinforcement learning through symbiotic evolution
TLDR
A new reinforcement learning method called SANE (Symbiotic, Adaptive Neuro-Evolution), which evolves a population of neurons through genetic algorithms to form a neural network capable of performing a task, is presented.
Learning fuzzy control by evolutionary and advantage reinforcements
In this paper, evolutionary and dynamic programming‐based reinforcement learning techniques are combined to form an unsupervised learning scheme for designing autonomous optimal fuzzy logic control
Evolving neural networks
TLDR
Methods that evolve fixed-topology networks, network topologies, and network construction processes, ways of combining traditional neural network learning algorithms with evolutionary methods, and applications of neuroevolution to game playing, robot control, resource optimization, and cognitive science are reviewed.
Evolving neural networks
TLDR
Methods that evolve fixed-topology networks, network topologies, and network construction processes, ways of combining traditional neural network learning algorithms with evolutionary methods, and applications of neuroevolution to game playing, robot control, resource optimization, and cognitive science are reviewed.
...
...

References

SHOWING 1-10 OF 51 REFERENCES
Designing Application-Specific Neural Networks Using the Genetic Algorithm
TLDR
A general and systematic method for neural network design based on the genetic algorithm, NeuroGENESYS, that employs the backpropagation learning rule and produces networks that perform significantly better than the randomly generated networks of its initial population.
Generalization and Scaling in Reinforcement Learning
TLDR
This paper describes a neural network algorithm called complementary reinforcement back-propagation (CRBP), and reports simulation results on problems designed to offer differing opportunities for generalization.
Is Learning Rate a Good Performance Criterion for Learning?
Evolving neural network controllers for unstable systems
  • A. Wieland
  • Mathematics
    IJCNN-91-Seattle International Joint Conference on Neural Networks
  • 1991
TLDR
The author describes how genetic algorithms were used to create recurrent neural networks to control a series of unstable systems, including network controllers with two, one, and zero inputs, and variations of the pole balancing problem.
Training Feedforward Neural Networks Using Genetic Algorithms
TLDR
A set of experiments performed on data from a sonar image classification problem are described to illustrate the improvements gained by using a genetic algorithm rather than backpropagation and chronicle the evolution of the performance of the genetic algorithm as it added more and more domain-specific knowledge into it.
Neuronlike adaptive elements that can solve difficult learning control problems
TLDR
It is shown how a system consisting of two neuronlike adaptive elements can solve a difficult learning control problem and the relation of this work to classical and instrumental conditioning in animal learning studies and its possible implications for research in the neurosciences.
Genetic algorithms and neural networks: optimizing connections and connectivity
Learning to control an inverted pendulum using neural networks
An inverted pendulum is simulated as a control task with the goal of learning to balance the pendulum with no a priori knowledge of the dynamics. In contrast to other applications of neural networks
Backpropagation and neurocontrol: a review and prospectus
  • P. Werbos
  • Psychology
    International 1989 Joint Conference on Neural Networks
  • 1989
The author reviews highlights of the emerging field of neurocontrol and describes how backpropagation could be used in a new generation of controllers suitable for problems involving thousands or
...
...