✨ Airline Passenger Satisfaction Predictor - Ai Project✨

An end-to-end Ai project that predicts airline passenger satisfaction using real-world travel survey data.It combines data preprocessing, machine learning, visualization, and a polished GUI using Tkinter.

---

🚀 Project Overview

This project aims to classify airline passengers as Satisfied or Neutral/Dissatisfied based on features related to their flight experience.It includes:

• 🔎 Exploratory Data Analysis (EDA) with Seaborn & Matplotlib

• 🧼 Data cleaning (handling missing values, encoding, outlier removal)

• 🔁 Feature scaling using StandardScaler

• 🧠 Model training using:

  • XGBoost (final model)
  • Logistic Regression
  • Random Forest
  • SVM
  • KNN
  • Decision Tree
  • Gradient Boosting
  • Neural Network (TensorFlow)

• 📈 Evaluation using Accuracy, Precision, Recall, F1 Score, ROC-AUC

• 🎯 Selection of best-performing model (XGBoost)

• 🖼️ Heatmaps and performance visualizations for comparison

---

🧩 Technologies Used

• Python (pandas, numpy, seaborn, matplotlib, scikit-learn, xgboost, tensorflow)
• Tkinter (for GUI)
• Jupyter Notebook (for analysis and development)

---

📦 How to Run

1. Clone the repo

   git clone https://github.com/Yomna-Mahsoob/Airline_Passanger_satisfication

2. Run to train and save the model:

   python airline_model.py

3. Launch the GUI:

   python GUI.py

---

📁 Files Structure

📂 project-folder
├── airline_model.py       # Full analysis, model training, evaluation
├── GUI.py          # GUI app for predictions
├── xgb_model.pkl             # Saved XGBoost model
├── scaler.pkl                # Saved StandardScaler

---

💡 Inspiration

This project is an example of turning a machine learning model into a user-friendly product.It's perfect for those learning how to bridge the gap between data science and real-world usability.

---

🎀 The GUI

The project features a custom-designed Tkinter GUI with:

• Pink-themed visual design 🎨
• Airplane icon integration ✈️
• Friendly input fields for passenger details
• One-click prediction using the trained XGBoost model
• Helpful pop-up messages for input errors and results

---

🖼️ GUI Preview

Here’s a quick preview of the interface:

---

🩷 Made BY:

• Thoraya Khaled
• Yomna EL-Kobesy
• Roquia Muhammad
• Roquia Rady
• Shahd Mazen
• Shrouk Bekheet
• Zainab Gamal