An Investigation into Predicting Flight Fares in India Using Machine Learning Models

Abstract

The number of passengers traveling on aeroplanes in India is increasing, and so are price changes. Seasonal and special event fluctuations occur in Indian airfares during different periods of the year. The challenge is to accurately predict flight prices. The research work presented in this paper has explored several machine learning models to predict airfares based on multiple characteristics which enhances the flight price prediction accuracy. This research proposes basic and advanced regression models that have been investigated seeking to accurately predict the price of the airline for a passenger to encourage passengers to make the flight ticket booking at the most optimal cost. The basic and advanced regression models explored in this research are Linear Regression, Decision Tree as a regressor, Random Forest as a regressor, XG Boost regressor, K-neighbours regressor, Bagging Regressor and Extra Trees regressor. The performance of these models was evaluated based on the Mean Average Error (MAE), Root Mean Square Error (RMSE), and adjusted R-Square metrics. After the evaluation of all the models that were implemented, Results analysis was discussed which showed that XG boost model achieved the best performance evaluation and the predicted prices were almost matching the actual price values of the flights.