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

Visualizing the function computed by a feedforward neural network.

T A Plate1, J Bert, J Grace

  • 1Bios Group LP, Santa Fe, NM 87501, USA.

Neural Computation
|August 10, 2000
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

Crust Composition and the Shallow Heat Source in KS 1731-260.

Physical review letters·2025
Same author

CT respiratory motion synthesis using joint supervised and adversarial learning.

Physics in medicine and biology·2024
Same author

Photosynthesis of tropical tree seedlings in relation to light and nutrient supply.

The New phytologist·2021
Same author

Radiomics analysis of 3D dose distributions to predict toxicity of radiotherapy for lung cancer.

Radiotherapy and oncology : journal of the European Society for Therapeutic Radiology and Oncology·2020
Same author

Improved Woodcock tracking on Monte Carlo simulations for medical applications.

Physics in medicine and biology·2018
Same author

Interactive effects of ultraviolet radiation and elevated CO<sub>2</sub> concentration on photosynthetic characteristics of European beech saplings during the vegetation season.

Plant physiology and biochemistry : PPB·2018
Same journal

A Model-Free Reinforcement Learning Implementation of Decision Making Under Uncertainty by Sequential Sampling.

Neural computation·2026
Same journal

DROP: Distributional and Regular Optimism and Pessimism for Reinforcement Learning.

Neural computation·2026
Same journal

Hierarchical Active Inference Using Successor Representations.

Neural computation·2026
Same journal

W-Kernel and Its Principal Space for Frequentist Evaluation of Bayesian Estimators.

Neural computation·2026
Same journal

A Hidden Markov Model-Inspired Sequence Classification Method for Hyperdimensional Computing.

Neural computation·2026
Same journal

Sparse Graphical Modeling for Electrophysiological Phase-Based Connectivity Using Circular Statistics.

Neural computation·2026
See all related articles

A new visualization method reveals how feedforward neural networks work, showing input effects and deviations from linearity. This technique aids in understanding complex models and identifying interactions for improved predictive accuracy.

Area of Science:

  • Machine Learning
  • Data Visualization
  • Statistical Modeling

Background:

  • Interpreting complex machine learning models like neural networks remains a significant challenge.
  • Understanding the input-output relationship is crucial for model validation and trust.
  • Existing methods often struggle with non-linearities and interactions within models.

Purpose of the Study:

  • To develop a novel visualization technique for understanding feedforward neural network functions.
  • To illustrate the impact of individual inputs and deviations from linearity.
  • To facilitate the identification of feature interactions in predictive models.

Main Methods:

  • A visualization method based on the input-output relationship of the model.
  • Applicable to any predictive statistical model, including ensemble methods.

Related Experiment Videos

  • Demonstrated on a neural network predicting lung cancer risk from smoking and drinking data.

    • Main Results:

    • The visualization clearly shows the effect of each input on the model's output.
    • Deviations from linear behavior are readily apparent.
    • The method successfully identified interactions within the lung cancer risk model.

    Conclusions:

    • The proposed visualization method enhances the interpretability of feedforward neural networks.
    • It is versatile, applicable to various predictive models beyond neural networks.
    • This approach aids in understanding complex biological relationships, such as those in cancer risk assessment.