Can Airline Passenger Satisfaction Be Predicted?

---

Overview

• Purpose: To put data from a Airline Customer Satisfaction Survey through various Machine Learning Models to best predict passenger customer satisfaction. Exploratory Data Analysis is performed to determine important factors in customer satisfaction. This helps the company better understand what goes into customer satisfaction in turn reducing customer churn.

• Data Source: Kaggle: Airline Passenger Satisfaction

   * The dataset contains satisfaction survey information from airline customers.
  

Questions We'd Like To Answer!

• Which 5 variables are most influential in customer satisfaction?

• Which machine learning model or deep learning model has a higher accuracy score when determining airline customer satisfaction?

• Tools Used:

  • Pandas Library, Numpy, SQLAlchemy, Scikitlearn, psycopg2, tensorflow, eli5
  • Visualization: Matplotlib, Seaborn

Models Used

We initally assigned the following models to each person:

Model Type	Team Member
DL Neural Networks	Irene
Decision Tree & Random Forest Classifier	Shanika
Naive Bayes Classifier	Vimal
Gradient Boosting Classifier	Wuyang

Dashboard Tool

• The team used Tableau & Google Slides to create our visual dashboard.

---