Picked a data set of about 104,000 different passenger surveys from an anonymous airline to determine ratings based on their experiences.
Analyzed different features relating to a customers flight experience to determine which features have the biggest impact on overall satisfaction. Some features that were analyzed were WiFi availability, Food & Drink service, Leg Room availability, and Inflight Entertainment options.
- Kaggle
The skills used to complete this project consisted of:
- Working with Python to make visualizations using Matplotlib & Seaborn
- Using Pandas to collect and clean the dataset
- Building & interpreting various classification models based on feature engineering/selection & hyperparameter tuning
- Using Scikit-Learn to compute various metrics relating to classification models
Two separate notebooks were posted on GitHub. One was the Final Project notebook which consisted of the Data Cleaning, Collection & Modeling components of the project and the ReadMe notebook which is a layout of how our project was presented.
- How much of an impact does Inflight Entertainment have on customer satisfaction?
- How much of an impact does Leg Room space have on customer satisfaction?
- How much of an impact does Inflight Wifi service have on customer satisfaction?
- How much of an impact does Food & Drink service have on customer satisfaction?
- What percent of customers were satisfied with their overall flight experience?
The data was gathered from about 104,000 different passenger surveys on an anonymous airline. After the data was gathered & cleaned, EDA was performed to see which features most strongly correlated to customer satisfaction. Following that, the data was split into training and testing sets and the resulting models were analyzed based on the different F-1 & Accuracy scores. Finally, the different models were compared to see which could best predict custome satisfaction.
- To analyze another customer satisfaction survey dataset from a different airline in order to confirm that customer opinions are similar amongst various airlines
- Potentially merge the two datasets together and apply more feature engineering/selection & hyperparameter tuning
Based on the results that were analyzed, the following recommendations can be made:
- Inflight Wifi, Entertainment, Food & Drink services and Leg Room space are very impactful on customer satisfaction.
- Customers prefer high speed WiFi internet connections. Overall it appeared that customers would rather not have wifi than have slow WiFi speed.
https://docs.google.com/presentation/d/1-GKvWRApTPEAVIxo0dNgPHWWjxHDjR8gD_uSmkumfg4/edit?usp=sharing
