Home / Current Issue / Paper 1717258
Flight Price Prediction Using Machine Learning and Deep Learning: A Comparative Study
Subject area: Science,Engineering and Technology · Area of research: Machine Learning
DOI: https://doi.org/10.64388/IREV9I11-1717258
Abstract
Airfare pricing is a highly dynamic and complex phenomenon influenced by numerous variables including departure time, number of stops, days to departure, flight class, and seasonal demand patterns. Accurate fare prediction offers practical value for cost-sensitive travelers and revenue-management optimization by airlines. This work presents a systematic comparative evaluation of eleven regression algorithms, spanning classical machine learning and contemporary deep learning approaches. Classical models include Linear Regression, Ridge, Lasso, Decision Tree, Random Forest, Extra Trees, Bagging, K-Nearest Neighbors, Gradient Boosting, and XGBoost. Five deep tabular architectures are benchmarked: MLP, DeepResNet1D, AttentionNet, WideAndDeep, and TabTransformer. Six CNN backbones (VGG11, VGG13, ResNet18, ResNet34, MobileNetV2, MobileNetV3) are also evaluated using synthetic 2-D image representations. All models are assessed across seven metrics: MAE, MSE, RMSE, R², Adjusted R², RMSLE, and MAPE. Results show that TabTransformer and ExtraTreesRegressor achieve R² exceeding 0.99.
Keywords
Airfare Price Prediction, Machine Learning, Deep Learning, Regression, Random Forest, XGBoost, TabTransformer, CNN, Comparative Study.
References
[1] O. Etzioni et al., "To Buy or Not to Buy: Mining Airfare Data to Minimize Ticket Purchase Price," Proc. ACM SIGKDD, pp. 119-128, 2003.
[2] W. Groves and M. Gini, "On Optimizing Airline Ticket Purchase Timing," ACM Trans. Intell. Syst. Technol., vol. 7, no. 1, 2015.
[3] K. Tziridis et al., "Airfare Prices Prediction Using ML Techniques," Proc. EUSIPCO, pp. 1036-1039, 2017.
[4] S. O. Arik and T. Pfister, "TabNet: Attentive Interpretable Tabular Learning," Proc. AAAI, vol. 35, pp. 6679-6687, 2021.
[5] X. Huang et al., "TabTransformer: Tabular Data Modeling Using Contextual Embeddings," arXiv:2012.06678, 2020.
[6] H. Cheng et al., "Wide & Deep Learning for Recommender Systems," Proc. DLRS Workshop at RecSys, 2016.
[7] T. Chen and C. Guestrin, "XGBoost: A Scalable Tree Boosting System," Proc. ACM SIGKDD, pp. 785-794, 2016.
[8] L. Breiman, "Random Forests," Machine Learning, vol. 45, no. 1, pp. 5-32, 2001.
[9] K. He et al., "Deep Residual Learning for Image Recognition," Proc. IEEE CVPR, pp. 770-778, 2016.
[10] A. Howard et al., "Searching for MobileNetV3," Proc. IEEE ICCV, pp. 1314-1324, 2019.
[11] K. Simonyan and A. Zisserman, "Very Deep Convolutional Networks," Proc. ICLR, 2015.
[12] F. Pedregosa et al., "Scikit-learn: Machine Learning in Python," JMLR, vol. 12, pp. 2825-2830, 2011.
[13] A. Paszke et al., "PyTorch: An Imperative Style Deep Learning Library," Proc. NeurIPS, vol. 32, 2019.
How to cite this paper
@article{1717258,
author = {Dhanush, Purna Satwik, Sandeep},
title = {Flight Price Prediction Using Machine Learning and Deep Learning: A Comparative Study},
journal = {Iconic Research And Engineering Journals},
year = {2026},
volume = {9},
number = {11},
pages = {115-121},
issn = {2456-8880},
url = {https://www.irejournals.com/formatedpaper/1717258.pdf},
abstract = {Airfare pricing is a highly dynamic and complex phenomenon influenced by numerous variables including departure time, number of stops, days to departure, flight class, and seasonal demand patterns. Accurate fare prediction offers practical value for cost-sensitive travelers and revenue-management optimization by airlines. This work presents a systematic comparative evaluation of eleven regression algorithms, spanning classical machine learning and contemporary deep learning approaches. Classical models include Linear Regression, Ridge, Lasso, Decision Tree, Random Forest, Extra Trees, Bagging, K-Nearest Neighbors, Gradient Boosting, and XGBoost. Five deep tabular architectures are benchmarked: MLP, DeepResNet1D, AttentionNet, WideAndDeep, and TabTransformer. Six CNN backbones (VGG11, VGG13, ResNet18, ResNet34, MobileNetV2, MobileNetV3) are also evaluated using synthetic 2-D image representations. All models are assessed across seven metrics: MAE, MSE, RMSE, R², Adjusted R², RMSLE, and MAPE. Results show that TabTransformer and ExtraTreesRegressor achieve R² exceeding 0.99.},
keywords = {Airfare Price Prediction, Machine Learning, Deep Learning, Regression, Random Forest, XGBoost, TabTransformer, CNN, Comparative Study.},
month = {May},
doi = {https://doi.org/10.64388/IREV9I11-1717258}
}