Jovian
Sign In
Learn practical skills, build real-world projects, and advance your career

Topic: Predicting Airline Passenger Satisfation with help of ML

With the help of various python based machine learning and data science tools we're going to analyse and train the data. Then we'll make an attempt to build a machine learning model based on our information. The ML model will help us to predict whether a passenger is satisfied with the airline service or not.

Steps we're going to follow-

  1. Defining Problem
  2. Briefing about Data
  3. Importing Data
  4. Data Exploration
  5. Data Summarisation
  6. Modelling
  7. Experimentation & Evaluation
  8. Conclusion & References

1. Defining Problem

Given the airline passenger satisfaction survey information about the passengers, can we predict whether they are satisfies or not with the airline services ? Also what factors are highly correlated to a satisfied (or dissatisfied) passenger?

2. Briefing about Data

2.1 What does this data contains?

The Airline Passenger Satisfaction folder comes with two csv files, train.csv and test.csv.

The datasets contains 23 different attributes about an airline passenger satisfaction survey. All these attributes gives us useful information about the passengers and how they rated the diffferent services of the flights.

What factors lead to customer satisfaction for an Airline?

2.2 Where do we get the data from?

We got this open dataset from Kaggle Datasets.

Link: https://www.kaggle.com/teejmahal20/airline-passenger-satisfaction

2.3 What information do we get from this data?

The datasets contains 23 different attributes which are -

  1. Gender: Gender of the passengers
    • Male : Male Passenger
    • Female : Female Passenger
  2. Customer Type: The customer type as per loyality and behaviour
    • Loyal Customer
    • Disloyal Customer
  3. Age: The actual age of the passengers
    • In Range
  4. Type of Travel: Purpose of the flight of the passengers
    • Personal Travel
    • Business Travel
  5. Class: Travel class in the plane of the passengers
    • Eco : Economy Class
    • Eco Plus : Premium Economy Class
    • Business : Business Class
  6. Flight distance: The flight distance of this journey
  7. Inflight wifi service: Satisfaction level of the inflight wifi service
    • 0 : Not Applicable
    • General Rating: 1-5
  8. Departure/Arrival time convenient: Satisfaction level of Departure/Arrival time convenient
    • 0 : Not Applicable
    • General Rating: 1-5
  9. Ease of Online booking: Satisfaction level of online booking
    • 0 : Not Applicable
    • General Rating: 1-5
  10. Gate location: Satisfaction level of Gate location
    • 0 : Not Applicable
    • General Rating: 1-5
  11. Food and drink: Satisfaction level of Food and drink
    • 0 : Not Applicable
    • General Rating: 1-5
  12. Online boarding: Satisfaction level of online boarding
    • 0 : Not Applicable
    • General Rating: 1-5
  13. Seat comfort: Satisfaction level of Seat comfort
    • 0 : Not Applicable
    • General Rating: 1-5
  14. Inflight entertainment: Satisfaction level of inflight entertainment
    • 0 : Not Applicable
    • General Rating: 1-5
  15. On-board service: Satisfaction level of On-board service
    • 0 : Not Applicable
    • General Rating: 1-5
  16. Leg room service: Satisfaction level of Leg room service
    • 0 : Not Applicable
    • General Rating: 1-5
  17. Baggage handling: Satisfaction level of baggage handling
    • 0 : Not Applicable
    • General Rating: 1-5
  18. Check-in service: Satisfaction level of Check-in service
    • 0 : Not Applicable
    • General Rating: 1-5
  19. Inflight service: Satisfaction level of inflight service
  20. Cleanliness: Satisfaction level of Cleanliness
    • 0 : Not Applicable
    • General Rating: 1-5
  21. Departure Delay in Minutes: Minutes delayed when departure
    • In Continous Range (Minutes)
  22. Arrival Delay in Minutes: Minutes delayed when Arrival
    • In Continous Range (Minutes)
  23. Satisfaction: Airline satisfaction level ( Prediction Target )
    • Satisfaction (1) : Overall satisfied with the services
    • Neutral or Dissatisfaction (0) : Overall unsatisfied with the services

2.4 Evaluation Criteria ?

If we can reach up to 95% accuracy at predicting whether or not a customer is satisfied with the airline services, then only we'll pursue for further application purpose.

  1. Confusion Matrix
  2. Classification Report
  3. ROC Curve and AUC Value
  4. Cross Validated Scores

2.5 Any another infomation ?

Modules we are going to use : Pandas, Numpy, Plotly, Matplotlib, Seaborn, Jovian, Joblib, XGBoost, Scikit-Learn.

# !pip install pandas, numpy, plotly, matplotlib, seaborn, jovian, joblib, xgboost, scikit-learn --q