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

Related Concept Videos

Quantifying Work

Quantifying Work

As a system undergoes a change, its internal energy can change, and energy can be transferred from the system to the surroundings, or from the surroundings to the system.

Modeling and Similitude

Modeling and Similitude

Scaled modeling is a fundamental technique in engineering, enabling the study of large and complex systems by creating smaller, manageable replicas that recreate critical characteristics of the original. In hydrology and civil infrastructure, for example, scaled models of dams help analyze water flow, turbulence, and pressure. This method allows for accurate predictions of real-world behavior within a controlled environment, significantly reducing the cost and time involved in full-scale...

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Multi-center validation of a machine learning model for early detection of monoclonal immunoglobulin-related disorders using routine laboratory data.

Frontiers in immunology·2026

Same author

Targeting androgen receptor transcriptionally represses VCP and enhances the efficacy of oncolytic-immunotherapy in hepatocellular carcinoma.

Pharmacological research·2026

Same author

Mitotic Cdc42 waves encode PI(3,4)P<sub>2</sub> signaling and Golgi morphological state to control spindle scaling.

Science advances·2026

Same author

Characterizing selection signatures in coding and noncoding regions of 14,886 cancer genomes.

Journal of genetics and genomics = Yi chuan xue bao·2026

Same author

Coordinated changes in microbiota features, short-chain fatty acids, and peripheral clocks accompany fructo-oligosaccharide-associated metabolic improvement.

Food & function·2026

Same author

[Impact of effective health information acquisition on hemophilia-related health literacy among caregivers of underage hemophilia patients].

Beijing da xue xue bao. Yi xue ban = Journal of Peking University. Health sciences·2026

Same journal

Dynamic analysis and reliable mechanical optimization application of ring HNN effected with a memristive neuron.

Neural networks : the official journal of the International Neural Network Society·2026

Same journal

DAFF-Net: A detection and search method for small-scale low surface brightness galaxies.

Neural networks : the official journal of the International Neural Network Society·2026

Same journal

Quasi-synchronization for complex networks with hybrid pinning intermittent control.

Neural networks : the official journal of the International Neural Network Society·2026

Same journal

Physics-encoded convolutional neural operators for parametric PDEs: A convergence-guaranteed framework via pre-computed kernel fields.

Neural networks : the official journal of the International Neural Network Society·2026

Same journal

Exploiting audio-visual modalities in videos: Object detection via multi-stage bilateral coupling network.

Neural networks : the official journal of the International Neural Network Society·2026

Same journal

Reliability-aware modality completion with cross-modal distillation for federated learning with missing modalities.

Neural networks : the official journal of the International Neural Network Society·2026

See all related articles

Search research articles

Related Experiment Video

Updated: Jun 9, 2025

Utilizing Electroencephalography Measurements for Comparison of Task-Specific Neural Efficiencies: Spatial Intelligence Tasks

Utilizing Electroencephalography Measurements for Comparison of Task-Specific Neural Efficiencies: Spatial Intelligence Tasks

Published on: August 9, 2016

Wasserstein task embedding for measuring task similarities.

Xinran Liu¹, Yikun Bai¹, Yuzhe Lu²

¹Computer Science Department, Vanderbilt University, 2201 W End Ave, Nashville, 37235, TN, United States.

Neural Networks : the Official Journal of the International Neural Network Society

|October 25, 2024

Summary

This summary is machine-generated.

This study introduces a novel, model-agnostic method for measuring task similarities in machine learning using optimal transport theory. This approach enables faster and more effective task comparisons, crucial for transfer and meta-learning applications.

Keywords:

Continual learning Dataset similarity Optimal transport Task embedding Transfer learning

More Related Videos

A Psychophysics Paradigm for the Collection and Analysis of Similarity Judgments

A Psychophysics Paradigm for the Collection and Analysis of Similarity Judgments

Published on: March 1, 2022

The Innovation Arena: A Method for Comparing Innovative Problem-Solving Across Groups

The Innovation Arena: A Method for Comparing Innovative Problem-Solving Across Groups

Published on: May 13, 2022

Related Experiment Videos

Last Updated: Jun 9, 2025

Utilizing Electroencephalography Measurements for Comparison of Task-Specific Neural Efficiencies: Spatial Intelligence Tasks

Utilizing Electroencephalography Measurements for Comparison of Task-Specific Neural Efficiencies: Spatial Intelligence Tasks

Published on: August 9, 2016

A Psychophysics Paradigm for the Collection and Analysis of Similarity Judgments

A Psychophysics Paradigm for the Collection and Analysis of Similarity Judgments

Published on: March 1, 2022

The Innovation Arena: A Method for Comparing Innovative Problem-Solving Across Groups

The Innovation Arena: A Method for Comparing Innovative Problem-Solving Across Groups

Published on: May 13, 2022

Area of Science:

Machine Learning
Computer Science
Data Science

Background:

Measuring task similarity is vital for machine learning tasks like transfer, multi-task, continual, and meta-learning.
Existing methods for task similarity assessment are often architecture-dependent, relying on pre-trained models or forward transfer proxies.
These limitations hinder efficient and generalizable task comparison across diverse machine learning paradigms.

Purpose of the Study:

To introduce a novel, model-agnostic, and training-free task embedding method for supervised classification.
To address the limitations of current architecture-dependent approaches for measuring task similarities.
To enable efficient handling of datasets with partially disjoint label sets.

Main Methods:

Leveraging optimal transport theory to define a new task embedding for supervised classification.
Employing multi-dimensional scaling for label embedding, followed by concatenation with dataset samples.
Defining dataset distance using the 2-Wasserstein distance between updated samples.
Utilizing a 2-Wasserstein embedding framework to map tasks into a vector space.

Main Results:

The proposed task embedding is model-agnostic and training-free, capable of handling disjoint label sets.
Task comparison using the novel embedding is significantly faster than existing methods like Optimal Transport Dataset Distance (OTDD).
Numerical experiments demonstrate statistically significant correlations between the proposed distance and forward/backward transfer across image recognition datasets.

Conclusions:

The developed 2-Wasserstein embedding framework provides an efficient and effective approach for measuring task similarities.
This model-agnostic method advances the field of machine learning by offering a more generalizable tool for task comparison.
The findings have implications for improving performance in transfer learning, multi-task learning, and meta-learning applications.