DDoS Attacks Analysis with Cyber Data Forensics using Weighted Logistic Regression and Random Forest

The internet has become an additional domain of warfare, alongside land, air, space, and sea, as technology has advanced. Cyberattacks can have a significant impact on both individual and national security, making the security of cyberspace a key concern for researchers. One prominent type of attack is the Distributed Denial of Service (DDoS) attack. DDoS attacks flood a targeted network or service with a high volume of traffic from multiple compromised systems, causing a denial of service for legitimate users. These attacks are a significant threat to cloud-based services. This paper presents a cyber forensics approach to analyze and classify these attacks by collecting benchmark network traffic datasets which fall under the Bigdata category. This is because DDoS attacks can generate a large amount of traffic in a short period of time. An exploratory data analysis approach in the Machine Learning Phase is proposed. Dask, an extensible parallel computing framework built on Python with machine learning, is used to uncover insights and understand patterns that may not be immediately obvious from the table. The proposed approach is evaluated using benchmark DDoS data and demonstrates its effectiveness in identifying attack characteristics and providing insight for defense. The results show the usefulness of cyber forensics in improving cloud security.