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

Receiver Operating Characteristic Plot

Receiver Operating Characteristic Plot

A ROC (Receiver Operating Characteristic) plot is a graphical tool used to assess the performance of a binary classification model by illustrating the trade-off between sensitivity (true positive rate) and specificity (false positive rate). By plotting sensitivity against 1 - specificity across various threshold settings, the ROC curve shows how well the model distinguishes between classes, with a curve closer to the top-left corner indicating a more accurate model. The area under the ROC curve...

You might also read

Related Articles

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

Sort by

Same author

The role of computational pathology in predicting expression of selected immunohistochemical markers.

Polish journal of pathology : official journal of the Polish Society of Pathologists·2026

Same author

Molecular and spatial differences in the tumor microenvironment of high-grade serous ovarian cancers with short versus long-term survival.

Gynecologic oncology·2026

Same author

Advances in digital health: Multimodal intelligence and translational impact.

Computer methods and programs in biomedicine·2025

Same author

AI-driven glomerular morphology quantification: a novel pipeline for assessing basement membrane thickness and podocyte foot process effacement in kidney diseases.

Computer methods and programs in biomedicine·2025

Same author

Efficient annotation bootstrapping for cell identification in follicular lymphoma.

Computer methods and programs in biomedicine·2025

Same author

Explainability and uncertainty: Two sides of the same coin for enhancing the interpretability of deep learning models in healthcare.

International journal of medical informatics·2025

Search research articles

Related Experiment Video

Updated: Jul 1, 2025

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model

Published on: August 16, 2020

Improving classification accuracy of fine-tuned CNN models: Impact of hyperparameter optimization.

Mikolaj Wojciuk¹, Zaneta Swiderska-Chadaj¹, Krzysztof Siwek¹

¹Faculty of Electrical Engineering, Warsaw University of Technology, Warsaw, Poland.

|March 11, 2024

Summary

Optimizing hyperparameters for convolutional neural networks (CNNs) using transfer learning can boost classification accuracy by up to 6%. Careful selection of hyperparameters and balanced data subsets are key for effective CNN training.

Keywords:

Deep learning Fine-tuning Hyperparameters Optimization

More Related Videos

Deep Neural Networks for Image-Based Dietary Assessment

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

Selecting Multiple Biomarker Subsets with Similarly Effective Binary Classification Performances

Selecting Multiple Biomarker Subsets with Similarly Effective Binary Classification Performances

Published on: October 11, 2018

Related Experiment Videos

Last Updated: Jul 1, 2025

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model

Published on: August 16, 2020

Deep Neural Networks for Image-Based Dietary Assessment

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

Selecting Multiple Biomarker Subsets with Similarly Effective Binary Classification Performances

Selecting Multiple Biomarker Subsets with Similarly Effective Binary Classification Performances

Published on: October 11, 2018

Area of Science:

Computer Science
Machine Learning
Artificial Intelligence

Background:

Convolutional Neural Networks (CNNs) are highly popular, necessitating efficient hyperparameter optimization.
Current methods often lack guidance on crucial hyperparameters, optimal ranges, or data subset usage for optimization.

Purpose of the Study:

Investigate the impact of different hyperparameter optimization methods and selections on fine-tuned CNN models.
Assess hyperparameter importance and optimal ranges for CNN training.
Explore using data subsets for hyperparameter optimization.

Main Methods:

Compared four hyperparameter optimization techniques: grid search, random search, Bayesian optimization, and Asynchronous Successive Halving Algorithm (ASHA).
Evaluated hyperparameter importance and identified optimal value ranges.
Assessed the feasibility of using data subsets for hyperparameter tuning.

Main Results:

Hyperparameter optimization improved CNN classification accuracy by up to 6%.
Class distribution balance in data subsets is critical for optimal hyperparameter selection.
Identified key hyperparameters and their effective value ranges for CNN training.

Conclusions:

Hyperparameter optimization significantly impacts CNN performance, especially in transfer learning scenarios.
The effectiveness of optimization is highly dependent on the specific task and dataset.
Using balanced data subsets for optimization is a viable and crucial strategy.