Scholarlink

<p>This paper introduces a novel framework for accurately predicting Super (a, d)-Edge Antimagic Total Labeling (SEATL) values in dynamic and structurally intricate networks using cutting-edge deep learning techniques. The framework leverages real-time data streams from domains such as biological processes, electrical systems, and communication infrastructures to model ever-evolving graphs. A Scaled Dot-Product Attention-based Graph Neural Network (GNN) is proposed to efficiently extract and learn complex topological patterns while adhering to antimagic labeling requirements. This deep learning pipeline integrates dynamic graph construction, feature derivation, model training, and SEATL value estimation. Comparative evaluations highlight the method’s superior predictive precision and lower computational cost relative to traditional antimagic and edge-magic labeling strategies. The results demonstrate that attention-augmented GNNs offer a robust, scalable, and adaptive solution for intelligent graph labeling in practical, real-time settings.</p> Graphical Abstract <p></p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Graph Neural Networks for Efficient SEATL Value Prediction in Real-Time Dynamic Networks

  • 期刊论文

  • 发表日期：

    2026年2月6日

查看全文
  • N. Muthuselvi,
  • T. Saratha Devi

摘要

This paper introduces a novel framework for accurately predicting Super (a, d)-Edge Antimagic Total Labeling (SEATL) values in dynamic and structurally intricate networks using cutting-edge deep learning techniques. The framework leverages real-time data streams from domains such as biological processes, electrical systems, and communication infrastructures to model ever-evolving graphs. A Scaled Dot-Product Attention-based Graph Neural Network (GNN) is proposed to efficiently extract and learn complex topological patterns while adhering to antimagic labeling requirements. This deep learning pipeline integrates dynamic graph construction, feature derivation, model training, and SEATL value estimation. Comparative evaluations highlight the method’s superior predictive precision and lower computational cost relative to traditional antimagic and edge-magic labeling strategies. The results demonstrate that attention-augmented GNNs offer a robust, scalable, and adaptive solution for intelligent graph labeling in practical, real-time settings.

Graphical Abstract