✈️ Airline Passenger Satisfaction Prediction

This machine learning project predicts whether an airline passenger is Satisfied or Not Satisfied based on in-flight experience and service metrics. It includes an interactive Gradio app where users can enter details and get real-time satisfaction predictions.

---

🧠 Objective

To apply multiple supervised learning algorithms and build a reliable model that can predict customer satisfaction using airline service data.

---

📂 Project Structure

• Data Preprocessing:

  • Dropped irrelevant columns (id, Unnamed: 0)
  • Encoded categorical variables using LabelEncoder
  • Scaled numerical features using StandardScaler

• Model Training:

  • Tested multiple classifiers:
    LogisticRegression, DecisionTreeClassifier, RandomForestClassifier,
    GradientBoostingClassifier, KNeighborsClassifier, SVC
  • Selected the best-performing model based on accuracy score

• Deployment:

  • Built a Gradio interface to accept user input and return prediction
  • Used pickle to save/load the trained model

---

📊 Features Used

• Demographics: Gender, Age, Customer Type
• Flight Experience: Class, Type of Travel, Flight Distance
• Services Rated:
  • Inflight Wifi, Food and Drink, Online Boarding, Seat Comfort,
    Entertainment, Cleanliness, Check-in Service, Leg Room Service, etc.
• Time Delays: Departure Delay, Arrival Delay

---

🔍 Visualizations

• Pie chart for satisfaction distribution
• Bar plots for average flight distance by class
• Heatmap of model confusion matrix for classification evaluation

📸 Screenshots

• 📊 Satisfaction Distribution Pie Chart
  

• 📉 Average Flight Distance by Travel Class
  

• 🧾 Confusion Matrix from Model Evaluation
  

• 🎛️ Gradio Interface
  

• 📈 Sample Prediction Output
  

---

🧪 ML Models Used

• Logistic Regression
• Decision Tree
• Random Forest
• K-Nearest Neighbors
• Gradient Boosting
• Support Vector Machine

✅ Final model chosen based on accuracy and generalization performance

---

🚀 How to Run the Project

• Clone the repo:

  git clone https://github.com/Anna-Ann11/passenger-satisfaction-prediction.git
  cd passenger-satisfaction-prediction
  

Install the required libraries:

pip install -r requirements.txt

Run the notebook:

jupyter notebook passenger_model.ipynb

Or launch the Gradio app:

iface.launch()