Proceedings ArticleDOI
Lamarckian Inheritance in Neuromodulated Multiobjective Evolutionary Neurocontrollers
Ian Showalter,Howard M. Schwartz +1 more
- pp 63-68
TLDR
It is shown that when Lamarckian inheritance is combined with evolved neurmodulated learning, neural controllers are synthesized in fewer generations than by neuromodulated evolution alone.Abstract:
This paper presents a novel evolutionary multiobjective neurocontroller with unsupervised learning and Lamarckian inheritance for robot navigation. Multiobjective evolution of network weights and topologies (NEAT-MODS) is augmented with Lamarckian inherited neuromodulated learning. NEAT-MODS is an NSGA-II augmented multiobjective neurocon-troller that uses two conflicting objectives. NEAT-MODS uses a selection process that aims to ensure Pareto-optimal genotypic diversity and elitism. Neuromodulation is a biologically-inspired technique that can adapt the per-connection learning rates of synaptic plasticity. Effectiveness of the design is demonstrated using a series of experiments with a simulated robot traversing a simple maze containing target goals. It is shown that when Lamarckian inheritance is combined with evolved neuromodulated learning, neural controllers are synthesized in fewer generations than by neuromodulated evolution alone. The proposed Lamarckian neuromodulated approach is found to be statistically superior to neuromodulation alone when applied to solve a multiobjective navigation problem.read more
Citations
More filters
Proceedings ArticleDOI
Multiobjective Neuromodulated Controllers for Efficient Autonomous Vehicles with Mass and Drag in the Pursuit-Evasion Game
Ian Showalter,Howard M. Schwartz +1 more
TL;DR: It is shown that compact and efficient neurocontrollers for pursuer agents with nonzero mass and drag, capable of capturing an optimal evader while simultaneously minimizing energy consumption, are evolved.
odNEAT: An Algorithm for Distributed Online, Onboard Evolution of Robot Behaviours
TL;DR: This work proposes and evaluates a novel approach to online distributed evolution of neural controllers called odNEAT, a completely distributed evolutionary algorithm for online learning in groups of embodied agents such as robots that approximates the performance of rtNEAT.
Proceedings ArticleDOI
Objective Comparison and Selection in Mono- and Multi-Objective Evolutionary Neurocontrollers
Ian Showalter,Howard M. Schwartz +1 more
TL;DR: In this paper, the effectiveness of individual elemental and compound objectives was compared to a mono-objective evolutionary neurocontroller. But the objective function selection was not directly compared, and it was shown that under certain circumstances, binary objectives can be unsuitable choices as objectives, and that it can be more effective to use multiobjective solutions than to combine elemental objective problems into monoobjective problems by auxiliary functions.
Proceedings ArticleDOI
Exploring the Relationship Between Topology and Function in Evolved Neural Networks
Ian Showalter,Howard M. Schwartz +1 more
TL;DR: In this article, the authors examine and analyze the topology of very small, minimally sized neurocontrollers that have been evolved for an extended number of generations, and demonstrate that patterns emerge in the neuromodulatory neurons, in the direct connections between neurocontroller inputs and outputs, and that topologies similar to those used in classical control are evolved.
References
More filters
Journal ArticleDOI
A fast and elitist multiobjective genetic algorithm: NSGA-II
TL;DR: This paper suggests a non-dominated sorting-based MOEA, called NSGA-II (Non-dominated Sorting Genetic Algorithm II), which alleviates all of the above three difficulties, and modify the definition of dominance in order to solve constrained multi-objective problems efficiently.
Book
Multi-Objective Optimization Using Evolutionary Algorithms
Kalyanmoy Deb,Deb Kalyanmoy +1 more
TL;DR: This text provides an excellent introduction to the use of evolutionary algorithms in multi-objective optimization, allowing use as a graduate course text or for self-study.
Book
The Organization of Behavior: A Neuropsychological Theory
TL;DR: In this paper, the authors discuss the first stage of perception: growth of the assembly, the phase sequence, and the problem of Motivational Drift, which is the line of attack.
Journal ArticleDOI
Evolving neural networks through augmenting topologies
TL;DR: Neural Evolution of Augmenting Topologies (NEAT) as mentioned in this paper employs a principled method of crossover of different topologies, protecting structural innovation using speciation, and incrementally growing from minimal structure.