FlightFarePredictionUsingMachineLearning

Predicting airline fares using machine learning regression models

This study enables customers to predict future airfares and also the airline industry to keep track of competitor prices on identical or comparable routes.For this study, over 300000 rows of airline data with around 10 key features for five Indian cities were collected and scraped from travel websites Expedia and Kayak. This data was used to perform exploratory data analysis followed by data cleaning and applying transformation techniques. The modeling was conducted using Machine Learning (ML) algorithms, Linear Regression, Random Forest, Decision Tree, K-Nearest Neighbours, and XGBoostRegressor. Machine Learning models showed significant results, which were helpful in predicting airfare. Among the implemented ML models, Random Forest and XGBoost gave the best results with over 98% accuracy. Using key elements like specific time periods of the day or the flight carrier, users can identify the best and cheapest flight.