The given dataset contains an airline passenger satisfaction survey. Our goal is to find what factors are highly correlated to a satisfied or an unsatisfied passenger? and perform predictions.
This project consists of : -
- Using sklearn Pipeline and ColumnTransformer for data transformation.
- Selecting the best ML algorithm based on size of the data and resources, and
- Hyperparameter tuning.
- Business class customers tend to be more satisfied.
- Business travel customers tend to be more satisfied.
- Customer satisfaction doesn't depend on Gender.
- Customers provided with services with 4-5 grade tend to be more satisfied.
- Cleanliness has high influence on satisfaction.
- All the provided services should be of 4-5 grade.
- Concentrate on Cleanliness.
- Concentrate on Eco class and personal travel as they are most unsatisfied customer.