Data transmission has skyrocketed with the ever-growing use of the Internet. And this data has constantly been an eye candy for the malicious intruders or attackers, and they are willing to obtain/alter it at any given point in time. And the growing frequency of these attacks has become a major hurdle in detecting intrusion as well as a threat to a system’s security over the Internet. An intrusion detection system is a software that monitors network traffic and sends alerts to the administrator. This paper puts forward a comparative study about how different machine learning techniques work with a high-dimensional dataset. This paper talks about the concept of intrusion detection system along with the taxonomy of an IDS in the first part and then different machine learning techniques explained used in the model. After that, methods used to evaluate an intrusion detection system are discussed, majorly discussing the results that were achieved by using various ml algorithms, emphasizing the benefits and drawbacks of each system. Throughout this research, experiments were conducted with CICIDS17 and CICIDS2018 datasets to construct the model. Lastly, it covers discussions regarding observations, future trends, and challenges along the research.