Airline companies are subject to a considerable amount of disruptions during their operations. It is vital for many industries including the airline industry to predict the sources of disruptions in different levels of management in order to reduce schedule recovery costs. One of the most important and costly sources of disruption in the airline industry is absenteeism of pilots at the time of a flight operation.

We propose a supervised learning method to predict total monthly absence hours of pilots. The proposed method uses characteristics of the monthly schedule as factors and, using an iterative algorithm, makes a prediction. The model was tested with real data and a substantial improvement was observed in the results.