Context and architecture
The analysis explores passenger and service records to identify patterns that are difficult to see through aggregate reporting alone.
Clustering, association rules, and dimensionality reduction were combined into one reproducible analysis workflow.
Core capabilities
- Passenger segmentation
- Association-rule mining
- Principal component analysis
- Operational insight reporting
Engineering challenge
Missing values and outliers required explicit cleaning before model fitting and comparison.
Outcome
Analyzed an estimated 50,000+ records and surfaced three actionable passenger-experience improvements.