Recently, the US air transport system has seen a pandemonium due to a computer breakdown. In the event of an aircraft delay being abruptly announced by the airport flight information display, people are all set to take off for their preferred destination. As a result of a system failure that provides safety information to pilots, thousands of flights across the country were canceled or delayed. The ongoing investigation into the malfunction, resulted in the hours-long grounding of many aircraft. To help better grasp the problem, we will explore in this study the many causes of why this occasionally occurs using machine learning techniques. There are five different types of delay that includes carrier, weather, NAS, security, and late arrival. Delay that is within the control of the National Airspace System (NAS) include parameters such as non-extreme weather conditions, airport operations, heavy traffic volume, air traffic control, etc. The nonparametric data visualization technique t-stochastic neighbor embedding (t-SNE) is used in classical machine learning. This is an unsupervised nonlinear technique mostly used for data exploration and high dimensional data visualization. The study will use t-SNE algorithm to classify and analyze NAS delay parameters and predict the area of high probability given by the actual and historical delay data. Results will be displayed in two-or three-dimensional space. t-SNE is a better technique than existing methods when it comes to producing a single map that depicts structure at numerous scales.