Gorde:
| Egile Nagusiak: | , , |
|---|---|
| Formatua: | Recurso digital |
| Hizkuntza: | ingelesa |
| Argitaratua: |
Zenodo
2025
|
| Gaiak: | |
| Sarrera elektronikoa: | https://doi.org/10.5281/zenodo.14890639 |
| Etiketak: |
Etiketa erantsi
Etiketarik gabe, Izan zaitez lehena erregistro honi etiketa jartzen!
|
Aurkibidea:
- <p><em><span>The rapid growth of the Internet of Things (IoT) has transformed industries, enabling seamless automation, smart homes, and interconnected industrial systems. However, this surge in connected devices has amplified security risks due to diverse communication protocols and limited computational resources in IoT devices. Traditional security measures often fall short in identifying complex and evolving threats. This research delves into the application of machine learning (ML) techniques for anomaly detection in IoT networks, aiming to spot irregular patterns that may indicate potential security breaches or malicious activities. By leveraging models such as decision trees, support vector machines, and neural networks, this study explores their efficacy in real-time and resource-constrained environments. The project also addresses challenges related to data quality, model optimization, and scalability. The anticipated results will contribute to more robust IoT security frameworks, providing an adaptive and intelligent approach to threat detection and prevention. Furthermore, this research highlights the importance of feature engineering and data preprocessing tailored for IoT environments, as the quality and variability of data play a crucial role in model performance. Comparative analyses will be conducted to evaluate the strengths and limitations of various ML algorithms, such as deep learning models like LSTM (Long Short-Term Memory) networks, which are well-suited for handling sequential and time- series data prevalent in IoT traffic. The study also incorporates evaluation metrics like detection accuracy, false positive rates, and computational efficiency to ensure the solutions are practical for deployment on edge devices with limited processing power. </span></em></p> <p><em><span> </span></em></p>