✈️ Airline Profit Prediction

Check out the Aviation Profit Predictor App

🔍 Ever wondered how airlines decide ticket prices, manage costs, and maximize profits?

This project unlocks the secrets behind airline profitability using cutting-edge machine learning and data-driven insights! 🚀

🎯 Project Overview

The Airline Profit Prediction system predicts airline profits based on real-world financial and operational metrics. It helps airlines make informed decisions, optimize resources, and boost revenue! 🔥

📊 Key Highlights

• 🧠 Smart Predictions: Uses historical airline data to predict future profits.
• 🛠️ Data Processing & Cleaning: Transforms messy data into meaningful insights.
• 💡 Feature Engineering: Creates powerful new metrics that improve accuracy.
• 📊 Power BI Dashboards: Interactive visualizations for deep analysis.
• 🔄 Automated Updates: Keeps predictions fresh and accurate.

🚀 Future Enhancements

We plan to integrate the following:

• 📡 API Deployment on Azure: Enable real-time profit prediction and easy access.
• 📊 Advanced Power BI Integration: Enhanced visualization with real-time analytics.
• 🔄 Continuous Model Improvement: Uploading new datasets to improve model accuracy.

🔬 Why Random Forest Regression Over Linear Regression?

We chose Random Forest Regression because:

• It captures non-linear relationships in airline data.
• It reduces overfitting by averaging multiple decision trees.
• It handles missing and categorical values better.

Linear Regression assumes a linear relationship, which may not hold in complex airline data.

⚖️ Why Standard Scaler Over MinMax Scaler?

We used Standard Scaler because:

• It ensures a zero mean and unit variance, making it suitable for ML models sensitive to feature scales.
• MinMax Scaler can distort data when outliers are present, making it less effective.

---

🚀 Quick Start Guide

Prerequisites

✔️ Python 3.8+
✔️ Virtual Environment
✔️ Power BI Desktop

Installation

# Clone the repository
git clone https://github.com/Sadnya20/Airline-Profit-Prediction.git
cd Airline-Profit-Prediction

# Create virtual environment
python -m venv venv
source venv/bin/activate   # On Windows use `venv\Scripts\activate`

# Install dependencies
pip install -r requirements.txt

---

🔥 How It Works

Train the Model

python main.py --mode train --data_path data/raw/airline_data.csv

Generate Predictions

python main.py --mode predict --input your_input.json

---

📊 Power BI Dashboard - See the Magic!

Want to explore your airline's profits like a pro? Our Power BI Dashboard turns raw data into beautiful, actionable insights! 🎨📈

1. Export Predictions: Model outputs are stored in data/processed/predictions.csv.
2. Load Data in Power BI: Import CSV and generate dynamic visual reports.
3. Automate Updates: Schedule refreshes for real-time insights.

PowerBi dashboard and Presentation

[https://drive.google.com/drive/folders/1LLH23USCa5rqFk09jE_nWsPHMv5TnJiI?usp=sharing]

---

💡 Contributors

• Sadnya Kolhe (Team Lead) - GitHub
• Daniyal Sheikh
• Sairaj Ajgaonkar
• Contributions welcome! 🎉

💙 What We Learned

This project taught us collaboration, problem-solving, and the fun of predictive modeling! 🚀