An enhanced deep learning model for phishing detection based on bayesian optimization and hybrid CNN VGG19 model

Abstract Phishing attacks, which trick users and obtain private data, are still a constant threat to cybersecurity. Using a large dataset of 10,000 samples, each with handcrafted features that capture URL structure and content-based cues, this study suggests an optimized deep learning framework for phishing website detection. The Binary Genetic Algorithm (BGA) is used for feature selection to improve model performance, and Bayesian Optimization (BO) is used for hyperparameter tuning. After normalization, the dataset is separated into subsets for training (70%), validation (15%), and testing (15%). CNN, VGG19, GRU, RNN, ResNet-50, and a CNN-VGG19 hybrid model are among the models that are assessed. According to experimental results, the CNN-VGG19 model performs best, achieving 0.9900 F1-score, 0.9901 precision, 0.99 recall, and 99.0% accuracy. This demonstrates how well convolutional neural networks and transfer learning work together to detect phishing attempts. The suggested model is a good fit for incorporation into real-time anti-phishing systems due to its robustness and strong classification capabilities. The combination of optimization, transfer learning, and deep feature extraction improves its performance.