Flight Price Prediction

Flight ticket prices fluctuate based on several factors, making price prediction a challenging task. This project applies machine learning techniques to analyze and predict ticket prices, helping travelers and businesses make informed decisions.

Project Sections

Overview

This project uses machine learning to predict flight ticket prices based on various features such as airline, departure and arrival times, duration, number of stops, and date of travel. The model is trained on a dataset containing flight pricing information and utilizes feature engineering techniques to improve prediction accuracy.

Features

  • Data preprocessing: Handling missing values, encoding categorical variables, and feature engineering
  • Feature selection and transformation for optimal model performance
  • Model training using Random Forest Regression for accurate price predictions
  • Hyperparameter tuning with GridSearchCV to optimize model performance
  • Performance evaluation using Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and R² Score
  • Visualization of feature importance to understand key factors affecting ticket prices

Technologies Used

Programming Language: Python

Libraries: Pandas, NumPy, Matplotlib, Scikit-learn

Machine Learning Model: Random Forest Regression

Project Media

Graphs

Image 1

This scatter plot visualizes the accuracy of the Random Forest Regression model by comparing the actual flight prices with the predicted prices. Ideally, points should lie along the diagonal line, indicating perfect predictions. The clustering along the diagonal suggests that the model performs well in estimating ticket prices, though some deviations are visible, indicating areas for potential improvement.

Image 2

This bar chart represents the importance of different features in predicting flight ticket prices. The "class" feature has the highest influence, followed by "duration", "days_left", and airline-related factors. This insight helps in understanding which variables significantly impact price fluctuations, guiding potential improvements to the model.

Repository Link

Explore the code and Data in the GitHub repository: GitHub - Flight Price Prediction