Computer networks currently provide a wide range of services to every organization. Banking, electric power, healthcare, telecommunications, and other critical network infrastructure are all supported by underlying computer networks. Furthermore, the digital revolution taking place in a variety of industries raises the relevance of computer networks. Computer networks, while crucial enablers of modern civilization, are not inherently harmful. However, numerous intrusions threaten network security and dependability, putting network-based services' confidentiality, integrity, or availability at risk. Intrusions may have significant consequences, such as denial of service (DoS), theft of critical data, and a single intrusion that can have a disastrous effect, causing the entire network infrastructure system to collapse. Intrusion Detection Systems (IDS) and Intrusion Prevention Systems (IPS) help to keep the network safe from intrusions. In recent years, hackers and cybercriminals have become increasingly able to employ complicated intrusion schemes to overcome traditional IDS/IPS technology. Traditional IDS/IPS solutions are vulnerable to erudite hackers and cybercriminals in developing nations. As a result, cybersecurity is rapidly becoming a top concern on the national security agenda. Thus, intelligent IDS/ Intrusion Detection and Prevention System (IDPS) capabilities that keep up with advanced threat monitoring, detection, and prevention are more vital than ever. To this end, Machine Learning-based classifiers are employed to enhance the intelligence of intrusion detection mechanisms. Previous studies show that more efforts have been given to single classifiers to develop and modernize IDSs, and less attention is given to hybrid classifier approaches. Thus, this research is oriented toward building and testing Machine Learning-based hybrid classifiers using the NSL-KDD dataset. After experimenting with various combinations of hybrid classifiers, this research has achieved the best performance using the hybrid of Naïve Bayes and Decision Tree (NBTree), obtaining an accuracy of 99.98%, Precision of 99.99%, Recall of 99.97%, and F1-Score of 99.98%. Based on the finding, this research concludes that the hybrid classifier approach is better than the single classifier approach and recommends that network infrastructure managers customize and upgrade their intrusion detection systems (IDSs) by employing a hybrid classifier. KEYWORDS: HYBRID CLASSIFIER, MACHINE LEARNING, DEEP LEARNING, DENIAL-OF-SERVICE, INTRUSION DETECTION SYSTEM