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

Real-time construction of optimized predictors from data streams.

Frank Kwasniok1, Leonard A Smith

  • 1Department of Statistics, London School of Economics, London, United Kingdom.

Physical Review Letters
|June 1, 2004
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

Semiparametric maximum likelihood probability density estimation.

PloS one·2021
Same author

A Four-Monoclonal Antibody Combination Potently Neutralizes Multiple Botulinum Neurotoxin Serotypes C and D.

Toxins·2021
Same author

Development of an o-pthalaldehyde (OPA) assay to measure protein content in Ricin Vaccine E. coli (RVEc™).

Vaccine·2020
Same author

Potency and stability of a trivalent, catalytically inactive vaccine against botulinum neurotoxin serotypes C, E and F (triCEF).

Toxicon : official journal of the International Society on Toxinology·2020
Same author

Detecting, anticipating, and predicting critical transitions in spatially extended systems.

Chaos (Woodbury, N.Y.)·2018
Same author

A Three Monoclonal Antibody Combination Potently Neutralizes Multiple Botulinum Neurotoxin Serotype E Subtypes.

Toxins·2018

This study introduces a novel method for building local models from data streams. It uses real-time updates to focus on important data regions for better system reconstruction.

Area of Science:

  • Machine Learning
  • Data Science
  • Dynamical Systems

Background:

  • Data streams present challenges due to their unlimited nature, making it impractical to use all observations for model building.
  • Traditional methods struggle to efficiently process and learn from continuous, high-volume data.
  • Accurate reconstruction of dynamical systems requires effective utilization of relevant data points.

Purpose of the Study:

  • To propose a new approach for constructing and optimizing local models specifically for data streams.
  • To address the challenge of impracticality in utilizing all observations from unlimited data sources.
  • To enhance the efficiency and accuracy of reconstructing underlying dynamical systems from streaming data.

Main Methods:

  • Developing a real-time revision mechanism for the learning set.

Related Experiment Videos

  • Implementing selective coverage of state space regions based on their contribution to system dynamics.
  • Utilizing local model construction and optimization techniques tailored for data stream environments.

    • Main Results:

    • Demonstrated the effectiveness of real-time learning set revision in managing data streams.
    • Showcased selective data coverage leading to improved reconstruction of dynamical systems.
    • Validated the proposed approach for local model construction and optimization in high-velocity data scenarios.

    Conclusions:

    • The novel approach offers an efficient way to build and optimize local models from data streams.
    • Real-time learning set revision is a key component for handling the scale and velocity of data streams.
    • This method significantly improves the ability to reconstruct underlying dynamical systems by focusing on critical data regions.