Network intrusion detection has become a critical component of modern cybersecurity systems due to the rapid growth of internet usage and the increasing volume and complexity of cyber-attacks. Cybersecurity protects computer systems and networks from unauthorized access [1]. This paper presents a comparative evaluation of three machine learning techniques namely Random Forest (RF), Decision Tree (DT), and K-Nearest Neighbors (KNN) for intrusion detection using the NSL-KDD benchmark dataset. The models were evaluated based on Accuracy, Precision, Recall, F1-score, Inference Time, and Confusion Matrix analysis. Experimental results show that Decision Tree achieved the fastest inference time (0.2397 sec) with balanced performance, while Random Forest achieved the highest precision (0.9677), indicating reliable attack detection with fewer false alarms. However, KNN, despite acceptable accuracy and precision, exhibited very high inference time (17.432366 sec), making it unsuitable for real-time deployment. The results highlight the importance of balancing detection performance and computational efficiency when selecting machine learning models for practical intrusion detection systems. These results demonstrate that Decision Tree provides an effective trade-off between detection accuracy and computational efficiency for real-time intrusion detection systems............... Keywords: ..............Machine learning, Intrusion Detection System, Decision Tree, Random Forest, KNN, NSL-KDD.