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

Confidence intervals for calibration with neural networks.

M Dathe1, M Otto

  • 1TU Bergakademie Freiberg, Institute for Analytical Chemistry, Leipziger Strasse 29, D-09599, Freiberg/Sa, Germany.

Analytical and Bioanalytical Chemistry
|August 1, 1996
PubMed
Summary

Neural network calibration using the bootstrap method accurately estimates confidence intervals for aromaticity determination in brown coals. This approach characterizes analysis errors and confirms the reliability of non-linear Back Propagation Neural Networks.

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

[Aerosol therapy in spontaneously breathing tracheostomized patients: an in vitro study].

Die Anaesthesiologie·2025
Same author

VGF AQEE- and GGEE-peptides differentiate between dementia types.

Journal of neurology·2025
Same author

Lipid-lowering and antioxidant properties of probiotic Bifidobacterium animalis MSMC83 in rats on a high-fat diet.

Beneficial microbes·2024
Same author

Frequency of an intrathecal IgM synthesis and MRZ reaction in children with MS.

European journal of paediatric neurology : EJPN : official journal of the European Paediatric Neurology Society·2024
Same author

[Cholecystolithiasis and intestinal bypass procedures].

Chirurgie (Heidelberg, Germany)·2023
Same author

[Massive pancarditis-autopsy report].

Pathologie (Heidelberg, Germany)·2023

Area of Science:

  • Analytical Chemistry
  • Spectroscopy
  • Machine Learning Applications

Background:

  • Accurate determination of aromaticity in brown coals is crucial for understanding their properties and potential applications.
  • Traditional methods for coal analysis may have limitations in precision and error characterization.
  • Infrared reflectance spectroscopy offers a potential route for rapid coal characterization.

Purpose of the Study:

  • To estimate confidence intervals for aromaticity determination in raw brown coals using neural network calibration.
  • To validate the efficacy of the bootstrap method for quantifying analysis errors in non-linear models.
  • To confirm the reliability of Back Propagation Neural Networks (BPNN) for spectroscopic analysis of coal.

Main Methods:

  • Development of a neural network calibration model based on infrared reflectance spectra of brown coals.

Related Experiment Videos

  • Application of the bootstrap method, a simplified Monte Carlo simulation, to estimate standard deviations and confidence intervals.
  • Utilizing Back Propagation Neural Networks (BPNN) as the non-linear analysis method for calibration.

    • Main Results:

    • Confidence intervals for aromaticity determination were successfully estimated using the bootstrap method.
    • The bootstrap method effectively characterized the analysis error associated with the neural network calibration.
    • The estimated confidence intervals provided statistical confirmation for the Back Propagation Neural Networks analysis.

    Conclusions:

    • The bootstrap method is a viable approach for estimating confidence intervals in non-linear calibration models like BPNN.
    • This methodology enhances the reliability assessment of aromaticity determination from infrared spectra of brown coals.
    • The study confirms the utility of BPNN coupled with bootstrap resampling for robust coal analysis.