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

Comments on ;Optimal training of thresholded linear correlation classifiers' [with reply].

D Lovell¹, A C Tsoi, T Downs

¹Dept. of Electr. Eng., Queensland Univ., St. Lucia, Qld.

IEEE Transactions on Neural Networks

|January 1, 1993

Summary

This summary is machine-generated.

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

Small beams, fast predictions: a comparison of machine learning dose prediction models for proton minibeam therapy.

Medical physics·2022

Same author

FIR and IIR Synapses, a New Neural Network Architecture for Time Series Modeling.

Neural computation·2019

Same author

Pursuing Quality in the Application of Bladder Cancer Quality of Life Research.

Bladder cancer (Amsterdam, Netherlands)·2016

Same author

Detection of tubule boundaries based on circular shortest path and polar-transformation of arbitrary shapes.

Journal of microscopy·2016

Same author

A comparison of upper and lower body energetics during high-intensity exercise.

The Journal of sports medicine and physical fitness·2015

Same author

Changes in sleep quantity and efficiency in professional rugby union players during home based training and match-play.

The Journal of sports medicine and physical fitness·2014

Same journal

Universal perceptron and DNA-like learning algorithm for binary neural networks: LSBF and PBF implementations.

IEEE transactions on neural networks·2013

Same journal

Guest editorial: special section on white box nonlinear prediction models.

IEEE transactions on neural networks·2011

Same journal

Data-based fault-tolerant control of high-speed trains with traction/braking notch nonlinearities and actuator failures.

IEEE transactions on neural networks·2011

Same journal

Guest editorial: special section on data-based control, modeling, and optimization.

IEEE transactions on neural networks·2011

Same journal

Neural network-based multiple robot simultaneous localization and mapping.

IEEE transactions on neural networks·2011

Same journal

Data-driven model-free adaptive control for a class of MIMO nonlinear discrete-time systems.

IEEE transactions on neural networks·2011

See all related articles

Researchers report issues with a neocognitron training algorithm where S-cells fail to detect trained features. This training vector rejection significantly impacts classification performance, a problem addressed by set theory explanations.

Area of Science:

Artificial Intelligence
Machine Learning
Computational Neuroscience

Background:

The neocognitron is a hierarchical, multilayered neural network inspired by the human visual cortex.
Closed-form training algorithms aim to simplify and accelerate the training process for such networks.
Previous work by T.H. Hildebrandt proposed a specific closed-form training algorithm for the neocognitron.

Purpose of the Study:

To report and analyze a difficulty encountered in the application of Hildebrandt's closed-form training algorithm for the neocognitron.
To investigate the impact of observed S-cell response failures on the neocognitron's classification performance.
To provide a response and explanation for the reported difficulties.

Main Methods:

Empirical testing of Hildebrandt's closed-form training algorithm on a neocognitron model.

Related Experiment Videos

Observation and documentation of S-cell response patterns during feature extraction.

Analysis of classification performance metrics in relation to training vector rejection.

Theoretical explanation based on set theory to address the observed issues.

Main Results:

Commenters observed that S-cells frequently failed to respond to features they were trained to extract.
Training vector rejection was identified as a significant factor affecting the neocognitron's classification performance.
Hildebrandt's reply suggests the observed negative results are not unique to the algorithm and can be explained by set theory principles.

Conclusions:

The application of Hildebrandt's closed-form training algorithm for neocognitrons may lead to S-cell response failures.
Training vector rejection is a critical issue impacting the performance of neocognitrons trained with this method.
The observed problems are potentially explainable through fundamental concepts in set theory, suggesting broader implications beyond the specific algorithm.