Heuristic optimization in classification atoms in molecules using GCN via uniform simulated annealing

Abstract Graph neural networks are becoming increasingly popular in deep learning due to their ability to process data in irregular structures and graphs thus preserving additional spatial dependencies due to the arrangement of nodes. This network allows for better accuracy results in classification...

Full description

Saved in:
Bibliographic Details
Main Authors: Agnieszka Polowczyk, Alicja Polowczyk, Marcin Woźniak
Format: Article
Language:English
Published: Nature Portfolio 2025-05-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-025-00340-8
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849731815905951744
author Agnieszka Polowczyk
Alicja Polowczyk
Marcin Woźniak
author_facet Agnieszka Polowczyk
Alicja Polowczyk
Marcin Woźniak
author_sort Agnieszka Polowczyk
collection DOAJ
description Abstract Graph neural networks are becoming increasingly popular in deep learning due to their ability to process data in irregular structures and graphs thus preserving additional spatial dependencies due to the arrangement of nodes. This network allows for better accuracy results in classification problems, where information from neighbors is also crucial. However, training such a model is complex and time-consuming and challenging. To date, some metaheuristic algorithms have been used primarily to optimize convolutional networks and find suitable hyperparameters, these are: genetic algorithm, particle swarm optimization, differential evolution or covariance matrix adaptation evolution strategy. In this paper, we propose a metaheuristic algorithm of Simulated Annealing with Uniform distribution for optimization of weights in GCNs, as a hybrid in combination with gradient optimizers. The performance of our technique was tested on the QM7 Dataset, where it was split into two datasets: imbalanced and balanced. Experimental results confirm that our proposed optimization method outperformed other standalone SOTA optimization models, including gradient and heuristics methods, demonstrating in each case to lower loss function values, higher accuracy values for balanced dataset and higher AUC (macro) values for imbalanced dataset.
format Article
id doaj-art-b62a984e8cdd472dabdf41f425c94723
institution DOAJ
issn 2045-2322
language English
publishDate 2025-05-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-b62a984e8cdd472dabdf41f425c947232025-08-20T03:08:25ZengNature PortfolioScientific Reports2045-23222025-05-0115111910.1038/s41598-025-00340-8Heuristic optimization in classification atoms in molecules using GCN via uniform simulated annealingAgnieszka Polowczyk0Alicja Polowczyk1Marcin Woźniak2Faculty of Applied Mathematics, Silesian University of TechnologyFaculty of Applied Mathematics, Silesian University of TechnologyFaculty of Applied Mathematics, Silesian University of TechnologyAbstract Graph neural networks are becoming increasingly popular in deep learning due to their ability to process data in irregular structures and graphs thus preserving additional spatial dependencies due to the arrangement of nodes. This network allows for better accuracy results in classification problems, where information from neighbors is also crucial. However, training such a model is complex and time-consuming and challenging. To date, some metaheuristic algorithms have been used primarily to optimize convolutional networks and find suitable hyperparameters, these are: genetic algorithm, particle swarm optimization, differential evolution or covariance matrix adaptation evolution strategy. In this paper, we propose a metaheuristic algorithm of Simulated Annealing with Uniform distribution for optimization of weights in GCNs, as a hybrid in combination with gradient optimizers. The performance of our technique was tested on the QM7 Dataset, where it was split into two datasets: imbalanced and balanced. Experimental results confirm that our proposed optimization method outperformed other standalone SOTA optimization models, including gradient and heuristics methods, demonstrating in each case to lower loss function values, higher accuracy values for balanced dataset and higher AUC (macro) values for imbalanced dataset.https://doi.org/10.1038/s41598-025-00340-8Graph convolutional networkSimulated annealingMetaheuristic optimizationUniform distributionNode classification
spellingShingle Agnieszka Polowczyk
Alicja Polowczyk
Marcin Woźniak
Heuristic optimization in classification atoms in molecules using GCN via uniform simulated annealing
Scientific Reports
Graph convolutional network
Simulated annealing
Metaheuristic optimization
Uniform distribution
Node classification
title Heuristic optimization in classification atoms in molecules using GCN via uniform simulated annealing
title_full Heuristic optimization in classification atoms in molecules using GCN via uniform simulated annealing
title_fullStr Heuristic optimization in classification atoms in molecules using GCN via uniform simulated annealing
title_full_unstemmed Heuristic optimization in classification atoms in molecules using GCN via uniform simulated annealing
title_short Heuristic optimization in classification atoms in molecules using GCN via uniform simulated annealing
title_sort heuristic optimization in classification atoms in molecules using gcn via uniform simulated annealing
topic Graph convolutional network
Simulated annealing
Metaheuristic optimization
Uniform distribution
Node classification
url https://doi.org/10.1038/s41598-025-00340-8
work_keys_str_mv AT agnieszkapolowczyk heuristicoptimizationinclassificationatomsinmoleculesusinggcnviauniformsimulatedannealing
AT alicjapolowczyk heuristicoptimizationinclassificationatomsinmoleculesusinggcnviauniformsimulatedannealing
AT marcinwozniak heuristicoptimizationinclassificationatomsinmoleculesusinggcnviauniformsimulatedannealing