The network attack detection frameworks are developed to find out the access to computing systems that are unauthorizedly connected across the networks. The intrusion detection is one of such frameworks, developed by that has a higher accuracy for all majority attacks in comparison to existing works. The models deploy different classifiers to demonstrate that the approach is modular in structure. Intrusion detection model developed in this analytical research utilises various machine learning classifiers like Random Forest, SVM, K-Nearest Neighbor, and Naïve Bayes. Experimentation was conducted on dataset NSLKDD, The Performance of classifiers improved as dimensionality reduction and feature selection improves accuracy and reduces false alarm rate. A better generalization is also achieved while integrating multiple classifiers. High accuracy is obtained for all majority attacks in the NSLKDD datasets which is the widely available benchmark datasets for intrusion detection.