Real‑Time Anomaly Detection Using Machine Learning: An Investigative Study

Chennapragada Amarendra, Kamineni Sai Kamal, Challa Kireeti Vardhan, Shaik Mohaseen, Belum Tirumala Reddy

2025

Abstract

As networked systems are continuously evolving, it is imperative to establish automated solutions for effective and efficient real-time anomaly detection for operational assurance and security. Traditional techniques can fall short in dynamic environments, leading to the adoption of machine learning (ML) methods. This paper investigates the performance of two models of ML in detecting anomalies in synthetic univariate time-series datasets simulating network traffic: ARIMA (a statistical method), and LSTM (a deep-learning model). Some advantages are, main metrics like accuracy or accuracy on learning, update to concept drift and computational cost. Findings highlight LSTM’s superior capability at handling those non-stationary data and seasonality challenges compared to ARIMA, although ARIMA is a possible path forward where resources are limited. We hope that the proposed RM-ML framework can demonstrate a significant potential for real-time monitoring of cyber threats in applications where ML can be utilized.

Paper Citation

in Harvard Style

Amarendra C., Kamal K., Vardhan C., Mohaseen S. and Reddy B. (2025). Real‑Time Anomaly Detection Using Machine Learning: An Investigative Study. In Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25; ISBN 978-989-758-777-1, SciTePress, pages 398-402. DOI: 10.5220/0013883700004919

in Bibtex Style

@conference{icrdicct`2525,
author={Chennapragada Amarendra and Kamineni Kamal and Challa Vardhan and Shaik Mohaseen and Belum Reddy},
title={Real‑Time Anomaly Detection Using Machine Learning: An Investigative Study},
booktitle={Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25},
year={2025},
pages={398-402},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013883700004919},
isbn={978-989-758-777-1},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25
TI - Real‑Time Anomaly Detection Using Machine Learning: An Investigative Study
SN - 978-989-758-777-1
AU - Amarendra C.
AU - Kamal K.
AU - Vardhan C.
AU - Mohaseen S.
AU - Reddy B.
PY - 2025
SP - 398
EP - 402
DO - 10.5220/0013883700004919
PB - SciTePress