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

PSECMAC: a novel self-organizing multiresolution associative memory architecture.

S D Teddy¹, C Quek, E K Lai

¹Centre for Computational Intelligence, Nanyang Technological University, Singapore 639798, Singapore.

IEEE Transactions on Neural Networks

|April 9, 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

In Vivo Image-Based 4D Modeling of Competent and Regurgitant Mitral Valve Dynamics.

Experimental mechanics·2021

Same author

A novel cognitive interpretation of breast cancer thermography with complementary learning fuzzy neural memory structure.

Expert systems with applications·2020

Same author

Financial volatility trading using a self-organising neural-fuzzy semantic network and option straddle-based approach.

Expert systems with applications·2020

Same author

PSECMAC intelligent insulin schedule for diabetic blood glucose management under nonmeal announcement.

IEEE transactions on neural networks·2010

Same author

DCT-Yager FNN: a novel Yager-based fuzzy neural network with the discrete clustering technique.

IEEE transactions on neural networks·2008

Same author

POPFNN-AAR(S): a pseudo outer-product based fuzzy neural network.

IEEE transactions on systems, man, and cybernetics. Part B, Cybernetics : a publication of the IEEE Systems, Man, and Cybernetics Society·2008

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

A new pseudo-self-evolving CMAC (PSECMAC) network improves on the standard CMAC model by adaptively allocating memory, inspired by cerebellar function. This novel approach enhances accuracy, memory use, and generalization for complex control tasks.

Area of Science:

Neuroscience and Computational Intelligence
Artificial Neural Networks
Control Systems Engineering

Background:

The cerebellar model articulation controller (CMAC) network, inspired by the cerebellum, is widely used but suffers from suboptimal accuracy, poor memory utilization, and the generalization-accuracy dilemma due to its regular structure.
Existing solutions to CMAC limitations often increase operational complexity.
The cerebellum's ability to model nonlinear dynamics is key, but its computational models face inherent challenges.

Purpose of the Study:

To introduce a novel neurophysiologically inspired associative memory architecture, the pseudo-self-evolving CMAC (PSECMAC) network.
To overcome the architectural deficiencies of the traditional CMAC network by employing nonuniform memory allocation.
To enhance modeling accuracy, memory utilization, and generalization capabilities.

Related Experiment Videos

Main Methods:

Developed the PSECMAC network with nonuniform memory allocation inspired by cerebellar synaptic plasticity.
Implemented a data-driven adaptive memory quantization scheme to emulate biological synaptic plasticity.
Utilized a neighborhood-based activation process for learning and computation.
Provided theoretical assurance of training stability through learning convergence proofs.

Main Results:

The PSECMAC network demonstrated significant improvements in memory utilization, output accuracy, and generalization capability compared to CMAC and its variants.
Benchmarking on currency futures option pricing, banking failure classification, and glucose-insulin dynamics confirmed PSECMAC's effectiveness.
The proposed nonuniform memory allocation scheme successfully addressed CMAC's architectural deficiencies.

Conclusions:

The PSECMAC network offers a superior alternative to traditional CMAC architectures for complex modeling and control tasks.
The neurophysiologically inspired design, particularly the adaptive memory allocation, is crucial for improved performance.
PSECMAC effectively balances accuracy, memory efficiency, and generalization, outperforming existing CMAC variants in real-world applications.