Ensemble Learning Techniques for Advanced Threat Detection in Complex Data Environments for Smart Education DOI

Virender Dhiman

Advances in educational technologies and instructional design book series, Journal Year: 2024, Volume and Issue: unknown, P. 191 - 206

Published: Dec. 20, 2024

As educational environments increasingly leverage digital technologies, they become more susceptible to a myriad of cyber threats. This chapter explores the application ensemble learning techniques for advanced threat detection in complex data environments, particularly within smart education frameworks. Ensemble learning, which combines multiple machine models enhance predictive performance, provides robust solution identifying and mitigating threats real-time. By analyzing diverse datasets from various illustrates how these can improve accuracy efficiency systems. Furthermore, it discusses integration methods with emerging technologies such as IoT, big analytics, AI create comprehensive security framework tailored education. Case studies demonstrating successful implementations highlight effectiveness adapting evolving landscape.

Language: Английский

Enhancing Intrusion Detection in Wireless Sensor Networks Using a GSWO-CatBoost Approach DOI Creative Commons
Thuan Minh Nguyen, Hanh Hong-Phuc Vo, Myungsik Yoo

et al.

Sensors, Journal Year: 2024, Volume and Issue: 24(11), P. 3339 - 3339

Published: May 23, 2024

Intrusion detection systems (IDSs) in wireless sensor networks (WSNs) rely heavily on effective feature selection (FS) for enhanced efficacy. This study proposes a novel approach called Genetic Sacrificial Whale Optimization (GSWO) to address the limitations of conventional methods. GSWO combines genetic algorithm (GA) and whale optimization algorithms (WOA) modified by applying new three-population division strategy with proposed conditional inherited choice (CIC) overcome premature convergence WOA. The achieves balance between exploration exploitation enhances global search abilities. Additionally, CatBoost model is employed classification, effectively handling categorical data complex patterns. A technique fine-tuning CatBoost’s hyperparameters introduced, using quantization strategy. Extensive experimentation various datasets demonstrates superiority GSWO-CatBoost, achieving higher accuracy rates WSN-DS, WSNBFSF, NSL-KDD, CICIDS2017 than existing approaches. comprehensive evaluations highlight real-time applicability method across diverse sources, including specialized WSN established benchmarks. Specifically, our GSWO-CatBoost has an inference time nearly 100 times faster deep learning methods while high 99.65%, 99.99%, 99.76%, 99.74% CICIDS2017, respectively.

Language: Английский

Citations

6
Enhancing Sound-Based Anomaly Detection Using Deep Denoising Autoencoder DOI Creative Commons
Seong‐Mok Kim,

Yong Soo Kim

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 84323 - 84332

Published: Jan. 1, 2024

Language: Английский

Citations

4
Toward Virtualized Optical-Wireless Heterogeneous Networks DOI Creative Commons
Zoran Vujičić, María C. Santos,

Rodrigo Méndez

et al.

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 87776 - 87806

Published: Jan. 1, 2024

Language: Английский

Citations

4
Variance‐driven security optimisation in industrial IoT sensors DOI Creative Commons
Hardik Gupta, Sunil K. Singh, Sudhakar Kumar

et al.

IET Networks, Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 16, 2024

Abstract The Industrial Internet of Things (IIoT) has transformed industrial operations with real‐time monitoring and control, enhancing efficiency productivity. However, this connectivity brings significant security challenges. This study addresses these challenges by identifying abnormal sensor data patterns using machine learning‐based anomaly detection models. proposed framework employs advanced algorithms to strengthen defences against cyber threats disruptions. Focusing on temperature anomalies, a critical yet often overlooked aspect security, research fills gap in the literature evaluating learning models for purpose. A novel variance‐based model is introduced, demonstrating high efficacy accuracy scores 0.92 0.82 NAB AnoML‐IOT datasets, respectively. Additionally, achieved F1 0.96 0.89 underscoring its effectiveness IIoT optimising cybersecurity processes. not only identifies vulnerabilities but also presents concrete solutions improve posture systems.

Language: Английский

Citations

4
Non-Intrusive Monitoring and Detection of Mobility Loss in Older Adults Using Binary Sensors DOI Creative Commons
Ioan Şuşnea, Emilia Pecheanu, Adina Cocu

et al.

Sensors, Journal Year: 2025, Volume and Issue: 25(9), P. 2755 - 2755

Published: April 26, 2025

(1) Background and objective: Mobility is crucial for healthy aging, its loss significantly impacts the quality of life, healthcare costs, mortality among older adults. Clinical mobility assessment methods, though precise, are resource-intensive economically impractical, most existing solutions automatic detection anomalies either obtrusive or improper long time monitoring. This study explores feasibility using non-intrusive, low-cost binary sensors continuous, remote in adults, aiming to identify both sudden events gradual loss. (2) Method: The utilized publicly available datasets (CASAS Aruba HH120) containing annotated activity data recorded from installed residential environments. After preprocessing-including filtering irrelevant sensor aggregation into behaviorally meaningful places (BMPs)-a series forecasting model (Prophet) was used predict normal patterns. A fuzzy inference module analyzed deviations between observed predicted determine probability anomalies. (3) Results: system effectively identified periods prolonged inactivity indicative potential falls other disruptions. Preliminary evaluation indicated a rate approximately 77-81% point anomalies, with false positive ranging 12 16%. Additionally, approach successfully detected simulated declines (1% per day reduction), evidenced by statistically significant regression trends levels over time. (4) Conclusions: argues that non-intrusive sensors, combined lightweight models inference, may provide practical scalable solution detecting Although performance can be further enhanced through improved preprocessing, predictive modeling, anomaly threshold tuning, proposed addresses key limitations approaches.

Language: Английский

Citations

0
Contextual Anomaly Detection in Smart Homes Using Temporal Graph Based Distances DOI
Amirhosein Bodaghi, Chris Nugent

Lecture notes in networks and systems, Journal Year: 2024, Volume and Issue: unknown, P. 118 - 128

Published: Jan. 1, 2024

Language: Английский

Citations

0
Ensemble Learning Techniques for Advanced Threat Detection in Complex Data Environments for Smart Education DOI

Virender Dhiman

Advances in educational technologies and instructional design book series, Journal Year: 2024, Volume and Issue: unknown, P. 191 - 206

Published: Dec. 20, 2024

As educational environments increasingly leverage digital technologies, they become more susceptible to a myriad of cyber threats. This chapter explores the application ensemble learning techniques for advanced threat detection in complex data environments, particularly within smart education frameworks. Ensemble learning, which combines multiple machine models enhance predictive performance, provides robust solution identifying and mitigating threats real-time. By analyzing diverse datasets from various illustrates how these can improve accuracy efficiency systems. Furthermore, it discusses integration methods with emerging technologies such as IoT, big analytics, AI create comprehensive security framework tailored education. Case studies demonstrating successful implementations highlight effectiveness adapting evolving landscape.

Language: Английский

Citations

0