Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Support vector echo-state machine for chaotic time-series prediction.

Zhiwei Shi1, Min Han

  • 1School of Electronic and Information Engineering, Dalian University of Technology, Liaoning 116023, China.

IEEE Transactions on Neural Networks
|March 28, 2007
PubMed
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

Formation Mechanism of Well-Ordered Densely Packed Nanoparticle Superlattices Deposited from Gas Phase on Template-Free Surfaces.

Nanoscale research letters·2021
Same author

Interleukin-7 Is Associated with Clinical and Pathological Activities in Immunoglobulin A Nephropathy and Protects the Renal Proximal Tubule Epithelium from Cellular Fibrosis.

Current medical science·2021
Same author

Utility of non-HDL-C in predicting proteinuria remission of idiopathic membranous nephropathy: a retrospective cohort study.

Lipids in health and disease·2021
Same author

Hot Oscillatory Pressing of Carbon Nanotube-Reinforced Copper Matrix Nanocomposite.

Nanomaterials (Basel, Switzerland)·2021
Same author

Two orders of magnitude extra SERS enhancement on silver nanoparticle-based substrate induced by laser irradiation in nitrogen ambient.

Spectrochimica acta. Part A, Molecular and biomolecular spectroscopy·2021
Same author

Single-step-fabricated disordered metasurfaces for enhanced light extraction from LEDs.

Light, science & applications·2021
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 method, Support Vector Echo-State Machines (SVESMs), uses a "reservoir trick" for accurate chaotic time-series prediction. This approach offers robust and efficient nonlinear time-series forecasting with promising results on benchmark and real-world data.

Area of Science:

  • Computational intelligence
  • Nonlinear dynamics
  • Machine learning

Background:

  • Chaotic time-series prediction is challenging due to inherent nonlinear dynamics.
  • Traditional methods like Support Vector Machines (SVMs) often struggle with complex temporal patterns.
  • Recurrent Neural Networks (RNNs) can model sequences but may face training difficulties.

Purpose of the Study:

  • To propose a novel chaotic time-series prediction method combining SVMs and echo-state mechanisms.
  • To introduce Support Vector Echo-State Machines (SVESMs) as an efficient alternative for nonlinear time-series analysis.
  • To leverage the strengths of structural risk minimization in a high-dimensional state space.

Main Methods:

  • Developed Support Vector Echo-State Machines (SVESMs) by replacing the 'kernel trick' with a 'reservoir trick'.

Related Experiment Videos

  • Implemented linear Support Vector Regression (SVR) within a high-dimensional 'reservoir' state space.
  • Utilized regularization operators and robust loss functions for generalization and robustness.

    • Main Results:

    • SVESMs demonstrated effectiveness on the benchmark Mackey-Glass time series.
    • The method showed promising prediction results on real-world data, including sunspots and Yellow River runoff.
    • SVESMs, as a type of RNN, possess a convex objective function ensuring global, optimal, and unique solutions.

    Conclusions:

    • SVESMs offer an efficient and robust approach for nonlinear chaotic time-series prediction.
    • The 'reservoir trick' effectively handles nonlinearity, providing advantages over traditional kernel methods.
    • The proposed method shows significant potential for applications in various scientific and engineering fields.