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

Distributed fault tolerance in optimal interpolative nets.

¹Department of Electrical Engineering, Cleveland State University, Cleveland, OH 44115, USA. d.simon@ieee.org

IEEE Transactions on Neural Networks

|February 6, 2008

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

Neutrophil apoptosis mediated by nicotinic acid receptors (GPR109A).

Cell death and differentiation·2007

Same author

Pharmacological inhibition of GSK-3 is not strictly correlated with a decrease in tyrosine phosphorylation of residues 216/279.

Journal of neuroscience research·2007

Same author

Type 2 diabetes mellitus: epidemiology, pathophysiology, unmet needs and therapeutical perspectives.

Diabetes & metabolism·2007

Same author

Density functional study of structural and electronic properties of small bimetallic silver-nickel clusters.

The journal of physical chemistry. A·2007

Same author

Increased lipopolysaccharide-induced tumour necrosis factor-alpha, interferon-gamma and interleukin-10 production in atopic dermatitis.

The British journal of dermatology·2007

Same author

Anaplasma phagocytophilum in dogs in Germany.

Zoonoses and public health·2007

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

This study enhances the optimal interpolative (OI) classification network algorithm for improved fault tolerance in neural networks. The new method distributes weights smoothly, increasing network resilience against hardware failures.

Area of Science:

Computer Science
Artificial Intelligence
Machine Learning

Background:

Fault tolerance is crucial for reliable hardware implementations of neural networks.
Conventional neural network training algorithms often neglect fault tolerance, leading to suboptimal weight distributions.
Existing fault tolerance methods frequently use unrealistic fault models, such as stuck neurons.

Purpose of the Study:

To extend the recursive training algorithm for optimal interpolative (OI) classification networks to incorporate distributed fault tolerance.
To address the nonoptimal weight distribution issue in conventional OI Net learning algorithms concerning fault tolerance.
To introduce a more realistic approach to fault tolerance in neural networks through weight distribution.

Main Methods:

Related Experiment Videos

The recursive training algorithm for optimal interpolative (OI) classification networks was modified.
The enhanced algorithm aims for a smooth distribution of network weights.
This smooth distribution leads to low weight salience and distributed computation.

Main Results:

Trained OI Nets using the new algorithm demonstrated increased fault tolerance.
The Iris classification problem was used as a benchmark to validate the results.
The proposed method shows a practical way to improve fault tolerance in neural network hardware.

Conclusions:

The extended OI Net training algorithm effectively improves distributed fault tolerance.
Smooth weight distribution is key to achieving realistic fault tolerance in neural networks.
This research offers a valuable contribution to the development of robust neural network hardware.