Using Data mining techniques to improve the security of Internet of Things network

نوع: Type: Thesis

مقطع: Segment: masters

عنوان: Title: Using Data mining techniques to improve the security of Internet of Things network

ارائه دهنده: Provider: Amir Mahdi Ashouri

اساتید راهنما: Supervisors: Dr. Vahid khoda karami

اساتید مشاور: Advisory Professors:

اساتید ممتحن یا داور: Examining professors or referees: Dr. Nafise Soleimani - Dr. Amirsaman Kheirkhah

زمان و تاریخ ارائه: Time and date of presentation: 2025

مکان ارائه: Place of presentation: 62

چکیده: Abstract: With the rapid growth of the Internet of Things (IoT) and the increasing number of connected devices, ensuring the security of these networks has become a critical challenge. Traditional intrusion detection systems (IDS) are no longer sufficient to meet the security demands of IoT networks due to their limitations in processing large volumes of data and recognizing complex patterns. Consequently, this study introduces an innovative framework to enhance the performance of IDS by utilizing data collected from various IoT platforms. The proposed framework combines deep learning techniques with metaheuristic optimization algorithms to extract and select effective features. In this framework, a convolutional neural network (CNN) is employed as the core feature extraction mechanism, which generates more relevant and lower-dimensional representations of input data. Subsequently, a novel feature selection mechanism based on the Manta Ray Foraging Optimization (MRFO) algorithm is proposed. This algorithm selects the most influential features from those extracted by the CNN. The proposed framework was evaluated on four standard datasets, demonstrating competitive classification performance compared to other known optimization methods. The hybrid approach, which leverages CNN and MRFO, significantly improves the performance of intrusion detection systems in IoT environments. This approach enhances the detection of cyberattacks by eliminating unnecessary features, selecting the most effective ones, and improving the overall system accuracy.

فایل: ّFile: Download فایل