Anomaly Detection to Network Instrusion:A Systematic Literature Review

In networking, anomaly detection is essential to preserving network infrastructure performance, security, and dependability. The increasing volume of network traffic and the dynamic and complicated character of contemporary networks pose serious problems for conventional anomaly detection techniques. The application of machine learning (ML) approaches to identify network traffic anomalies, such as malicious activity, performance deterioration, and configuration problems, is examined in this study. Supervised, unsupervised, and semi-supervised machine learning models provide improved accuracy in detecting trends and anomalies in real-time data streams. The study compares the efficacy of many machine learning techniques, including decision trees, support vector machines, k-means clustering, and neural networks, in identifying various anomalies across a range of network contexts. The difficulties in handling imbalanced datasets, training models with labeled data, and lowering false positives are also highlighted in the research. It also talks about how anomaly detection systems can be integrated with intrusion detection systems and network monitoring tools to create a stronger security architecture. All things considered, using machine learning to networking anomaly detection is a promising strategy for risk reduction and guaranteeing the seamless operation of contemporary networked systems.