Optimizing Hyperparameters in Meta-Learning for Enhanced Image Classification

This paper investigates the significance of hyperparameter optimization in meta-learning for image classification tasks. Despite advancements in deep learning, real-time image classification applications often suffer from data inadequacy. Few-shot learning addresses this challenge by enabling learni...

Full description

Saved in:
Bibliographic Details
Main Authors: Amala Mary Vincent, P. Jidesh, A. A. Bini
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11087560/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850078821095571456
author Amala Mary Vincent
P. Jidesh
A. A. Bini
author_facet Amala Mary Vincent
P. Jidesh
A. A. Bini
author_sort Amala Mary Vincent
collection DOAJ
description This paper investigates the significance of hyperparameter optimization in meta-learning for image classification tasks. Despite advancements in deep learning, real-time image classification applications often suffer from data inadequacy. Few-shot learning addresses this challenge by enabling learning from limited samples. Meta-learning, a prominent tool for few-shot learning, learns across multiple classification tasks. We explore different types of meta-learners, with a particular focus on metric-based models. We analyze the potential of hyperparameter optimization techniques, specifically Bayesian optimization and its variants, to enhance the performance of these models. Experimental results on the Omniglot and ImageNet datasets demonstrate that incorporating Bayesian optimization, particularly its evolutionary strategy variant, into meta-learning frameworks leads to improved accuracy compared to settings without hyperparameter optimization. Here, we show that by optimizing hyperparameters for individual tasks rather than using a uniform setting, we achieve notable gains in model performance, underscoring the importance of tailored hyperparameter configurations in meta-learning.
format Article
id doaj-art-801db74569154c0b8fd411c070d4ff50
institution DOAJ
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-801db74569154c0b8fd411c070d4ff502025-08-20T02:45:28ZengIEEEIEEE Access2169-35362025-01-011313081613083110.1109/ACCESS.2025.359114211087560Optimizing Hyperparameters in Meta-Learning for Enhanced Image ClassificationAmala Mary Vincent0https://orcid.org/0000-0002-0009-5017P. Jidesh1https://orcid.org/0000-0001-9448-1906A. A. Bini2https://orcid.org/0000-0002-0559-267XDepartment of Mathematical and Computational Sciences, National Institute of Technology Karnataka, Surathkal, Mangalore, IndiaDepartment of Mathematical and Computational Sciences, National Institute of Technology Karnataka, Surathkal, Mangalore, IndiaDepartment of Electronics and Communication Engineering, National Institute of Technology Karnataka, Surathkal, Mangalore, IndiaThis paper investigates the significance of hyperparameter optimization in meta-learning for image classification tasks. Despite advancements in deep learning, real-time image classification applications often suffer from data inadequacy. Few-shot learning addresses this challenge by enabling learning from limited samples. Meta-learning, a prominent tool for few-shot learning, learns across multiple classification tasks. We explore different types of meta-learners, with a particular focus on metric-based models. We analyze the potential of hyperparameter optimization techniques, specifically Bayesian optimization and its variants, to enhance the performance of these models. Experimental results on the Omniglot and ImageNet datasets demonstrate that incorporating Bayesian optimization, particularly its evolutionary strategy variant, into meta-learning frameworks leads to improved accuracy compared to settings without hyperparameter optimization. Here, we show that by optimizing hyperparameters for individual tasks rather than using a uniform setting, we achieve notable gains in model performance, underscoring the importance of tailored hyperparameter configurations in meta-learning.https://ieeexplore.ieee.org/document/11087560/Meta-learningfew-shot learningimage classificationhyperparameter optimizationevolutionary algorithms
spellingShingle Amala Mary Vincent
P. Jidesh
A. A. Bini
Optimizing Hyperparameters in Meta-Learning for Enhanced Image Classification
IEEE Access
Meta-learning
few-shot learning
image classification
hyperparameter optimization
evolutionary algorithms
title Optimizing Hyperparameters in Meta-Learning for Enhanced Image Classification
title_full Optimizing Hyperparameters in Meta-Learning for Enhanced Image Classification
title_fullStr Optimizing Hyperparameters in Meta-Learning for Enhanced Image Classification
title_full_unstemmed Optimizing Hyperparameters in Meta-Learning for Enhanced Image Classification
title_short Optimizing Hyperparameters in Meta-Learning for Enhanced Image Classification
title_sort optimizing hyperparameters in meta learning for enhanced image classification
topic Meta-learning
few-shot learning
image classification
hyperparameter optimization
evolutionary algorithms
url https://ieeexplore.ieee.org/document/11087560/
work_keys_str_mv AT amalamaryvincent optimizinghyperparametersinmetalearningforenhancedimageclassification
AT pjidesh optimizinghyperparametersinmetalearningforenhancedimageclassification
AT aabini optimizinghyperparametersinmetalearningforenhancedimageclassification