A Multi-Model Machine Learning Framework for Robust Anomaly-Based Threat Detection

Abstract

Cybersecurity is essential in the contemporary, rapidly evolving digital environment. As AI-driven solutions become essential for protecting businesses, the increasing number and complexity of cyber threats often surpass traditional security measures, leading to considerable financial and reputational risks. This paper presents an enhanced cybersecurity framework utilizing an ensemble learning model that integrates machine learning and deep learning methods to tackle this difficulty. We assessed and contrasted various classifiers utilizing the HIKARI-2021 dataset from Kaggle, including Random Forest, Decision Tree, Gaussian Naive Bayes, K-Nearest Neighbors, Logistic Regression, Multi-Layer Perceptron, and Convolutional Neural Network. By merging these models using an ensemble technique, we harnessed their complementary capabilities, attaining a remarkable 96.32% accuracy—an appreciable enhancement above individual models. In addition to precision, the ensemble method improves adaptability, facilitating more dynamic and robust security frameworks. Our research underscores the effectiveness of ensemble learning in cybersecurity, illustrating its capacity to strengthen digital companies against emerging threats. This research provides practical solutions and facilitates further studies on AI-integrated cybersecurity, promoting innovation and a resilient global digital infrastructure.