Problem Solving with a Perpetual Evolutionary Learning Architecture

Authors:
Jong-Chen Chen
Affiliations:
Department of Management Information Systems, National YunLin University of Science and Technology, Touliu, Taiwan
Venue:
Applied Intelligence
Year:
1998

Citing 5
Cited 1

Multifacetted modelling and discrete event simulation

Multifacetted modelling and discrete event simulation
A multilevel neuromolecular architecture that uses the extradimensional bypass principle to facilitate evolutionary learning

Proceedings of the NATO advanced research workshop and EGS topical workshop on Chaotic advection, tracer dynamics and turbulent dispersion
Evolutionary computation: toward a new philosophy of machine intelligence

Evolutionary computation: toward a new philosophy of machine intelligence
Pattern categorization and generalization with a virtual neuromolecular architecture

Neural Networks
Data Structures and Algorithms

Data Structures and Algorithms

Review: Neuromolecularware and its application to pattern recognition

Expert Systems with Applications: An International Journal

Quantified Score

Hi-index	0.00

Visualization

Abstract

The capability of learning in an indefinite amount of time rendersbiological systems highly adaptable. We have developed a biologicallymotivated computer model, called the artificial neuromolecular (ANM)system, that demonstrates long-term evolutionary learning capability forcomplex problem solving. The major elements of the system are neurons whoseinput-output behavior is controlled by significant internal dynamics. Thedynamics are modeled by cellular automata, structured to represent theneuronal cytoskeleton (a subneuronal network found in every neuron).Neurons of this type are linked into a multilayer network that abstractssome features of visual circuitry. Multiple copies of these networks arecontrolled by neurons with memory manipulation capabilities. The ANM systemcombines these two types of neurons into a single, closely integratedarchitecture. The system is educated to perform desired tasks byevolutionary algorithms. These algorithms act at the intraneuronal level togenerate a repertoire of neurons with different pattern processingcapabilities. They also act at the interneuronal level (through the memorymanipulation system) to orchestrate different pattern processing neuronsinto a group suitable for performing desired tasks. The system has beenapplied to Chinese character recognition. Experiments were emphasized on long-term evolutionary learning, relearning capability, self-organizing dynamics, malleability, gradual transformability, multidimensional fitnesssurface, co-evolutionary learning, and cross-level synergy.