In the descriptive analytics section, we omitted the origin airport, however, in the predictive section, we used the origin airport as one of the main independent variables that appeared to have the lowest MAPE, providing us with a better combination for accuracy of flight delays. Other variables that were omitted include number of passengers, route code, destination airport, departure date, scheduled departure time, scheduled arrival time, delay indicator, and actual arrival time. We omitted those variables because they were not relevant for prediction purposes. If we included all three versions of times like scheduled departure and arrival time, they would create a collinearity. Furthermore, the number of passengers were eliminated because we did not have data or number of passengers for MDA flights, whereas, we only had data for RegionEx, which was not going to be an accurate way for comparison between the both airlines.