Scholarlink

Image spam detection presents significant challenges in ensuring the integrity of communication platforms. Advanced feature extraction techniques with EfficientNetB0 and supervised learning models, specifically Support Vector Machine (SVM), are employed and optimized through single-objective Bayesian optimization (BO) for hyperparameter tuning, aiming to maximize accuracy. The methodology incorporates nested cross validation to ensure robust model evaluation. Empirical results highlight the effectiveness of the approach, achieving state-of-the-art performance in spam image classification. The findings demonstrate the potential of this single-objective optimization framework for enhancing image spam detection.

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

Single-Objective Optimization for Image Spam Detection Using Supervised Learning and Bayesian Hyperparameter Tuning

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Adnane Filali,
  • El Arbi Abdellaoui Alaoui,
  • Mostafa Merras

摘要

Image spam detection presents significant challenges in ensuring the integrity of communication platforms. Advanced feature extraction techniques with EfficientNetB0 and supervised learning models, specifically Support Vector Machine (SVM), are employed and optimized through single-objective Bayesian optimization (BO) for hyperparameter tuning, aiming to maximize accuracy. The methodology incorporates nested cross validation to ensure robust model evaluation. Empirical results highlight the effectiveness of the approach, achieving state-of-the-art performance in spam image classification. The findings demonstrate the potential of this single-objective optimization framework for enhancing image spam detection.