• Kokkala Rachana
• Katroth BalaKrishna Maruthiram

Network security relies on a robust Intrusion Detection System (IDS) to swiftly identify and thwart passive and aggressive attacks. Leveraging machine learning, particularly on the NSL-KDD datasets, enhances the efficiency of anomaly detection. By employing a hybrid feature extraction method, combining various mathematical techniques, our model optimally processes network traffic data. This involves selecting pertinent features for training binary classification models, enabling accurate prediction of normal or attack types. The evaluation of our model's performance includes assessing overall accuracy and error rates, reinforcing the role of network traffic analysis in exposing invasions and preventing attacks. The project preprocesses the NSL-KDD dataset to extract relevant features and trains ML models for binary classification of normal behavior and attack patterns. Various ML algorithms are evaluated to determine the most effective intrusion detection approach. The ensemble methods which are Voting Classifier(RF + AdaBoost) and Stacking Classifier(LGBM + MLP + RF + XGB) are implemented, resulting in a noteworthy 99% accuracy in prediction.

Network Attack , Intrusion detection system, Machine learning, Attack types, Feature extraction

"MACHINE LEARNING SAFEGUARDS : NETWORK ATTACK DETECTION", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 8, page no.e832-e837, August-2024, Available :http://www.jetir.org/papers/JETIR2408488.pdf

"MACHINE LEARNING SAFEGUARDS : NETWORK ATTACK DETECTION", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 8, page no. ppe832-e837, August-2024, Available at : http://www.jetir.org/papers/JETIR2408488.pdf