Unleashing the Power of Deep Learning: Revolutionizing Cyber Threat Detection Systems

This paper discusses the need for new methodologies in cyber threat detection systems due to the increasing volume and speed of data generated by future technologies. The paper presents a deep learning-based method for detecting cyber attacks in wireless sensor networks (WSNs). The approach takes into account WSN node operations and MQTT data transport capabilities. By using a deep learning model instead of a regular machine learning model, the accuracy of cyber threat detection is improved. The paper emphasizes the importance of using deep learning models that can analyze network stream data to identify cyber threats, which can help prevent potential attacks and minimize the risk of system failure. Overall, the paper highlights the potential of deep learning-based cyber threat detection systems in addressing the challenges posed by the evolving landscape of cyber security.

Rughani et al. (2021) investigated the potential of machine learning techniques for detecting cyberbullying on social media platforms. The study collected data from Twitter and used various machine learning algorithms to classify tweets as cyberbullying or non-cyberbullying. The results showed that machine learning algorithms could accurately classify cyberbullying with high accuracy, sensitivity, and specificity. The study also highlighted the importance of considering various factors such as context and language in cyberbullying detection.

In another study, Huang et al. (2018) explored the use of machine learning in digital forensic investigations, specifically in the detection of steganography, which is the practice of concealing data within other data. The study compared the performance of various machine learning algorithms in detecting steganography and found that deep learning models, such as convolutional neural networks, had the highest accuracy. The study also highlighted the potential of machine learning in automating the process of steganography detection, which can be time-consuming and labor-intensive for human investigators.

Moreover, Oladipo et al. (2020) investigated the application of machine learning in digital forensics for object recognition, which is crucial in analyzing video surveillance footage and forensic images. The study focused on the use of convolutional neural networks for object identification and classification and highlighted its effectiveness in digital forensics. The study also identified six areas for improvement in future digital forensic studies.

In the field of network security, Ilyas and Alharbi (2022) explored the use of machine learning algorithms for network intrusion detection using the CSE-CICIDS2018 dataset. The study found that machine learning classifiers could effectively monitor for malicious activity on computer networks and highlighted the importance of rethinking and redesigning intrusion detection systems to adapt to the evolving nature of internet threats.

Finally, Goni and Mohammad (2020) presented a machine learning-based mobile forensics system to tackle cybercrime in Nigeria. The system utilized intelligent media towers and satellites to categorize calls as threats and relay notifications to the Nigerian Communication Commission’s forensic lab. The study emphasized the importance of using machine learning approaches to handle the large amounts of data generated by the Internet of Things (IoT) and the rise in mobile devices, which traditional data mining approaches may not be able to handle effectively.

Reference: https://easychair.org/publications/preprint/DL2D