Airline Ticket Price and Demand Prediction

Abstract

Air ticket price prediction has become a pivotal research area within data analytics and machine learning. With the growth of airline networks and online booking platforms, navigating dynamic pricing to secure affordable tickets remains a major challenge for passengers. This paper presents a structured framework for predicting flight ticket prices using machine learning regression algorithms and deep learning techniques. Using a benchmark Kaggle dataset, categorical features were preprocessed using Label Encoding, and various predictive models including Linear Regression, Decision Trees, and Random Forest ensemble methods were trained and evaluated. We analyze key internal and external factors driving fare volatility and assess the performance metrics of each algorithm. The optimal regression model was selected based on predictive accuracy, offering a robust and scalable approach to dynamic flight fare estimation.

Introduction

1. Microbially Induced Calcium Carbonate Precipitation (MICP) for Self-Healing Concrete

Concrete is widely used but prone to cracking, which reduces durability and increases maintenance costs. MICP-based self-healing concrete uses bacteria (mainly Bacillus species) to precipitate calcium carbonate that seals cracks, restores strength, reduces permeability, and extends service life. Recent studies focus on improving bacterial survival through encapsulation, immobilization, and endospore technologies, as well as predictive modeling and carbon-neutral construction. MICP also has applications in soil stabilization, environmental remediation, heavy metal immobilization, and restoration of historical structures. Despite its potential, challenges such as long-term bacterial viability, field-scale implementation, and standardized evaluation remain, highlighting the need for further research.

2. Transport-Oriented Planning (TOP) in Butuan City

This study evaluates Butuan City's readiness for Transport-Oriented Planning (TOP), which integrates transportation and land use to promote sustainable urban development. Using documentary analysis of planning documents such as the CLUP, Transportation Master Plan, and route plans, the study assessed land-use integration, accessibility, multimodal transport, sustainable mobility, and institutional support. Results indicate moderate to high readiness, with strong policy foundations but limited integration between land-use and transportation planning. Major gaps include insufficient active transport infrastructure, fragmented institutional coordination, and suburban expansion. The study recommends stronger policy integration, coordinated planning, and greater emphasis on walkability and cycling infrastructure.

3. Artificial Intelligence in ESG and Sustainable Finance

This review examines how Artificial Intelligence enhances Environmental, Social, and Governance (ESG) practices in sustainable finance using secondary data. AI technologies—including Natural Language Processing, Machine Learning, and Computer Vision—improve ESG data collection, greenwashing detection, climate risk assessment, and sustainable investment decisions. Evidence shows AI enables faster ESG controversy detection, better prediction of ESG scores, and improved risk-adjusted portfolio performance. However, challenges such as algorithmic bias, lack of explainability, data privacy, and inconsistent ESG standards remain. The paper proposes a four-layer Responsible AI framework incorporating data ingestion, AI processing, validation, and decision-making to promote transparent and reliable ESG integration.

4. DNA Barcoding for Species Identification

DNA barcoding is a molecular technique that accurately identifies species using standardized genetic markers, overcoming the limitations of traditional morphology-based taxonomy. Animals primarily use the mitochondrial COI gene, plants use rbcL and matK, while fungi rely on the ITS region. The study evaluated amplification success, species discrimination, and barcode gap analysis across multiple taxonomic groups. Results showed excellent performance for animals, particularly fishes and insects, while plants and fungi exhibited lower species-level discrimination due to evolutionary complexities. DNA barcoding has broad applications in biodiversity conservation, wildlife forensics, agriculture, medicinal plant authentication, and environmental DNA (eDNA) monitoring, making it an essential tool for modern biodiversity assessment.

5. CNN-Based Lip Reading System

This study develops a real-time automated lip-reading system using deep learning. The system captures lip movements from webcam video, detects the mouth region using OpenCV and dlib, and classifies spoken words using a Convolutional Neural Network (CNN). CNN performance was compared with LSTM and Temporal Convolutional Network (TCN) models, with CNN selected for deployment due to its superior speed and accuracy. Trained on a subset of the LRW dataset containing ten words, the model achieved 100% training accuracy and high-confidence real-time predictions. Although recognition accuracy is limited by dataset size and visually similar words, the lightweight CNN architecture demonstrated practical, low-latency performance suitable for interactive applications.

6. Airline Ticket Price Prediction Using Machine Learning

This study presents a machine learning-based system for predicting airline ticket prices and passenger demand using historical flight data. The proposed framework employs a Random Forest Regressor trained on features such as airline, origin, destination, travel class, distance, and days before departure. Data preprocessing includes cleaning, label encoding, feature selection, and an 85:15 train-test split. Model performance is evaluated using R² and Mean Absolute Error (MAE), with Random Forest outperforming Linear Regression and Decision Trees in prediction accuracy. The trained model is integrated into a Flask web application that provides real-time fare predictions, helping travelers optimize booking decisions while supporting airline revenue management.

Conclusion

In conclusion, machine learning techniques can be effectively used for airfare price prediction. By analyzing historical data, features such as flight route, date, time, airline, and various other factors can be used to train machine learning models to predict future airfare prices. These models can provide valuable insights and assist both airlines and passengers in making informed decisions. 1) Accurate Predictions: Machine learning models can analyze large volumes of historical data and identify patterns and trends that humans might miss. This allows for more accurate predictions of airfare prices, helping both airlines and passengers plan their travel budgets effectively. 2) Pricing Strategy Optimization: Airlines can utilize machine learning models to optimize their pricing strategies. By considering various factors such as demand, seasonality, competition, and customer behavior, airlines can adjust their fares in real-time, maximizing their revenue and profitability. 3) Personalized Pricing: Machine learning can enable personalized pricing based on individual customer preferences and behaviors. This can help airlines tailor offers to specific customer segments, enhancing customer satisfaction and loyalty. 4) Demand Forecasting: Machine learning models can assist in predicting future demand for flights. By analyzing historical booking data and external factors like events, holidays, and economic indicators, airlines can optimize flight schedules and capacity planning. 5) Dynamic Pricing: Machine learning algorithms can support dynamic pricing, allowing fares to be adjusted in real-time based on supply and demand dynamics. This can benefit airlines and passengers, ensuring optimal pricing and maximizing seat occupancy.

References

[1] P. Malighetti, S. Paleari and R. Redondi, \"Pricing strategies of low-cost airlines: The Ryanair case study\", Journal of Air Transport Management, vol. 15, no. 4, pp. 195-203, 2009. [2] P. Malighetti, S. Paleari and R. Redondi, \"Has Ryanair\'s pricing strategy changed over time? An empirical analysis of its 2006-2007 flights\", Tourism Management, vol. 31, no. 1, pp. 36-44, 2010. [3] W. Groves and M. Gini, \"A regression model for predicting optimal purchase timing for airline tickets\", Technical Report 11-025, 2011. [4] W. Groves and M. Gini, \"An agent for optimizing airline ticket purchasing\", 12th International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2013), pp. 1341-1342, May 06-10, 2013. [5] M. Papadakis, Predicting Airfare Prices, 2014. T. Janssen, A linear quantile mixed regression model for prediction of airline ticket prices, 2014. [6] R. Ren, Y. Yang and S. Yuan, Prediction of airline ticket price, 2015. [7] Hüseyin Korkmaz (2024) Prediction of Airline Ticket Price Using Machine Learning Method Journal of Transportation and Logistics, Vol. 9, No. 2, pp. 205–218, 2024. DOI: 10.26650/JTL.2024.1486696. This paper compares multiple ML algorithms for airline fare prediction and identifies key factors affecting ticket prices. [8] Yilong Wang, Shengrun Zhang, Xiaowei Tang, Chongheng Zhang (2025) Airline Ticket Price Prediction Based on CNN-LSTM Hybrid Model Journal of Beijing Jiaotong University, Vol. 48, Issue 5. Published January 2025. This study uses deep learning (CNN-LSTM) for airfare forecasting. [9] CheniSruneethi& I. Udaykiran (2025) Flight Price Prediction Using Machine Learning International Journal of Scientific Research in Computer Science, Engineering and Information Technology, May 2025. [10] SK Fhysuddin, Pallam Tejasri, Avula Swapna Sri, et al. (2025)Intelligent Fare Dynamics Modeling Using Machine Learning for Adaptive Airline Pricing ForecastsInternational Journal of Data Science and IoT Management System, Vol. 4, No. 2, 2025. The paper compares Linear Regression, Decision Tree, and Random Forest models for ticket price prediction. [11] Muzaffer Ertürk, Murat Emeç, Ay?e At?lgan Sar?do?an, Nabi Küçükgergerli (2025)Large-Scale Airline Ticket Price Prediction Using Ensemble Machine Learning AlgorithmsNi?anta??ÜniversitesiSosyalBilimler Dergisi, June 2025. Uses a dataset of over 300,000 Indian flight records and ensemble ML techniques.

Copyright

Copyright © 2026 Sahithi Varma Gottimukkala , Pravitha R Prasad . This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.