Airline Fare Prediction using Machine Learning

Project Overview

This project aims to predict airline fares using a Random Forest machine learning model. The model demonstrates strong predictive accuracy, achieving an R² score of 0.804. This predictive capability can assist travelers and businesses in making informed decisions regarding flight pricing.

Table of Contents

• Installation
• Usage
• Data Preprocessing
• Model Development
• Results
• Technologies Used
• License

Installation

To run this project, you need to have Python 3.6 or later installed. You can set up your environment by installing the required packages using pip:

pip install pandas numpy scikit-learn

Usage

1. Clone the repository: git clone https://github.com/SamJoeSilvano/Airline_Ticket_Fare_Prediction.git

2. Navigate to the project directory: cd airline-fare-prediction

3. Run the main script: python main.py

Data Preprocessing

The dataset is preprocessed to ensure high quality and accuracy in predictions. Key steps include:

• Data cleaning to remove any missing or inconsistent entries.
• Feature engineering, including feature selection and encoding of categorical variables.
• Splitting the data into training and testing sets.

Model Development

The model development involves:

• Using the Random Forest algorithm for regression.
• Hyperparameter tuning with RandomizedSearchCV to optimize model performance.
• Evaluating the model using metrics such as R² score and Mean Absolute Error (MAE).

Results

The Random Forest model achieved an R² score of 0.804, indicating a strong predictive capability for airline fare predictions.

Technologies Used

• Python
• Scikit-learn
• Pandas
• NumPy
• Jupyter Notebook (for exploratory analysis and visualization)

License

This project is licensed under the MIT License. See the LICENSE file for more details.