The detection of Distributed Denial-of-Service (DDoS) network attacks is critical for ensuring the security of online systems with increasing cyber threats. The study purposes advanced machine learning techniques such as XGBoost, Random Forest, and Logistic Regressor models to accurately detect DDoS attacks. XGBoost is an iterative approach, learning from residuals of preceding models to continuously enhance predictive capabilities. It aims to advance intrusion detection methods and strengthen cybersecurity measures with the surge in cyber-attacks affecting sectors like cryptocurrency companies and healthcare infrastructure. Traditional defense methods struggle to match the rising complexity of DDoS attack strategies while various machine and deep learning algorithms such as DBSCAN, SVM, Random Forest, modified CNNs, and DNNs for is used for DDoS detection. The XGBoost model is implemented after data preprocessing and feature selection of CICIDS2017 by using Python and Jupyter Notebook. The model is evaluated using precision, recall, F1-score, confusion matrices, and graphical representations such as precision-recall curve, roc-curve, and det-curve. The XGBoost model’s exhibits exceptional performance of precision, recall, and F1-score for both DDoS and non-DDoS classes. It achieves precision and recall scores of 0.99 for both classes with an accuracy of 1.00 while the Random Forest and Logistic Regressor models present unreliable efficiency. The study proposed XGBoost for reliable DDoS attack detection, and highlights its essential role in strengthening cybersecurity techniques and provides valuable insights for professionals.