<![CDATA[Distributed Denial of Service (DDoS) attacks have emerged as one of the most critical threats to contemporary network security. Rapid and accurate detection of such attacks is major for ensuring service continuity in large-scale networks. This study proposes an integrated approach that combines feature engineering with machine learning algorithms for the detection of DDoS attacks. In the initial phase, ANOVA and Chi-Square tests were applied to the dataset to identify statistically significant features; attributes such as dt, switch, dur, bytecount, and pktcount, which contributed minimally to classification performance or contained redundant information, were excluded. The optimized feature set was then evaluated using several machine learning algorithms, namely Decision Tree, Support Vector Machine (SVM), K-Nearest Neighbors (KNN), and Logistic Regression (LR). Quantitatively, feature selection improved SVM accuracy from 74.88% to 95.05%, increased Decision Tree accuracy to nearly 99.94%, slightly reduced KNN performance while maintaining its overall strength, and decreased LR accuracy from 77.15% to 74.87%. The experimental findings demonstrate that the proposed approach not only enhances classification performance but also reduces model runtime. Accordingly, the study presents an effective solution that simultaneously delivers high accuracy and computational efficiency in DDoS detection.]]>
Copyrights © 2025
