I calculated the health scores for 500 different cities based upon 7 different health metrics. After scoring each city, I was able to rank the healthiest and least healthy cities. I used both unsupervised learning and supervised learning to determine the ranks of the cities. I carried out a variety of tests to determine the best model for ranking.
This took me to create 5 different clustering models and 7 different classification models and each of their resulting pairs to determine the best model. I found that the KMeans clustering method and the LightGBM classification provided the best model for ranking cities.