A Novel Cuckoo Search-Based Optimized Deep CNN Model for Phishing Attack Detection in IoT Environment

Abstract

There is a serious risk to information privacy and security in the IoT ecosystem from phishing attacks. This research offers a unique strategy for increasing phishing attack detection by employing the Cuckoo Search algorithm to tune hyperparameters in a deep Convolutional Neural Network (CNN) model. Our suggested approach is very effective, with a 90\% success rate in detecting phishing attacks. To distinguish between legal and fraudulent IoT network activity, we systematically tune hyperparameters using Cuckoo Search to optimise the deep CNN model's performance. This study helps ensure the safety of sensitive information and IoT devices by providing a novel and effective approach for real-time phishing detection.