A Study on an Effective Model for Predicting Flight Delay

Abstract

Amongst the most significant business concerns that airline companies face is the considerable expenses related to airlines being delays caused due to natural events and operations and maintenance flaws, which is an additional expense for the airlines, having caused scheduling and operations problems for end-users, likely to result in a negative revenue and customer displeasure. We used supervised machine learning approaches in this study to develop a two-stage prediction models for forecasting flight on-time performance. This model's initial stage uses binary classification to predict flight delays, while the second phase uses regression to estimate the delay's duration in minutes. The proposed research compares the effectiveness of decision tree classifier to logistic regression. Based on the created model, the outcomes of this simulation disclose projected congestion in airports, considering hour, day, climate, and so on. As a result, there will be less time spent waiting.