Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Videos

Cooperative coevolution of neural representations.

A D Brown¹, H C Card

¹Department of Electrical and Computer Engineering, University of Manitoba, Winnipeg, Canada.

International Journal of Neural Systems

|October 29, 2000

Summary

A genetic algorithm (GA) optimizes feature detectors for multilayer perceptrons (MLPs) in image classification. This approach enhances network accuracy and confidence by evolving effective data representations.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Metrology of infrared superconducting bolometers with a backshort.

The Review of scientific instruments·2023

Same author

Excess Heat Capacity in Mo/Au Transition Edge Sensor Bolometric Detectors.

IEEE transactions on applied superconductivity : a publication of the IEEE Superconductivity Committee·2021

Same author

Shear behavior of human skull bones.

Journal of the mechanical behavior of biomedical materials·2021

Same author

DNA methylation across the genome in aged human skeletal muscle tissue and muscle-derived cells: the role of HOX genes and physical activity.

Scientific reports·2020

Same author

Natural Language Processing of Radiology Reports in Patients With Hepatocellular Carcinoma to Predict Radiology Resource Utilization.

Journal of the American College of Radiology : JACR·2019

Same author

The mechanical response of commercially available bone simulants for quasi-static and dynamic loading.

Journal of the mechanical behavior of biomedical materials·2018

Area of Science:

Computer Science
Artificial Intelligence
Machine Learning

Background:

Multilayer perceptrons (MLPs) are widely used for image classification.
Feature detection is crucial for effective data representation in neural networks.
Genetic algorithms (GAs) offer a powerful optimization framework for complex search spaces.

Purpose of the Study:

To investigate the use of genetic algorithms (GAs) for optimizing local feature detectors in multilayer perceptrons (MLPs).
To explore different encoding strategies for feature detectors within the GA's chromosome structure.
To evaluate the impact of these strategies on image classification accuracy and network confidence.

Main Methods:

A genetic algorithm (GA) was employed to search for optimal local feature detectors (hidden units).
Three distinct encoding methods for hidden unit weights were implemented: single chromosome coevolution and two cooperative approaches with individual chromosomes.
The fitness function incorporated both MLP classification accuracy and network confidence.

Main Results:

The study successfully utilized a GA to evolve feature detectors for MLP-based image classification.
Different encoding strategies demonstrated varying effects on the cooperation and performance of feature detectors.
The fitness function effectively guided the GA towards solutions balancing accuracy and confidence.

Conclusions:

Genetic algorithms provide an effective method for optimizing feature detectors in MLPs for image classification tasks.
Cooperative encoding strategies show promise in promoting the development of specialized and effective feature detectors.
The proposed approach offers a novel way to enhance neural network performance by integrating evolutionary computation with deep learning architectures.

Related Experiment Videos