Project 06
Choosing where to live in Calgary involves weighing safety, density, economic vitality, and housing mix across 200+ communities. This project applies unsupervised machine learning to group communities into distinct livability segments for residents, planners, and policymakers.
Interactive app
Interactive Calgary map with communities color-coded by livability segment assignment
Detailed breakdown of each segment showing average crime, population, and business metrics
Side-by-side comparison of any two communities across all 10 livability features
Multi-dimensional radar visualization of community feature profiles and cluster centroids
Performance
Approach
Integrated four Calgary Open Data sources: census, crime, business licences, and building permits. Built a 10-feature community-level matrix covering population, crime rate, business diversity, and more. Applied KMeans (k=2..10) and Agglomerative clustering with silhouette score selection. Reduced dimensionality with PCA for visualization and component interpretation.