The Food Establishment Inspection Scores data from the past three years has been provided by the city. The establishments were inspected on different dates, and the number of inspections varied across establishments, with some inspected more frequently than others. This report utilizes a clustering algorithm, guided by the elbow optimization method, to classify the establishments into four categories based on their average sanitation scores over the last three years.Using the last two inspection scores for establishments that were inspected at least twice, we found that only 44.4% showed improvement from their previous inspection.
264 of the total 6668 establishments received an average score of 100 over the past three years. These 264 establishments were inspected with varying frequencies. 151 were inspected only once which makes their scores less reliable compared to those with a higher number of inspections. Further details can be found in the report, along with four downloadable data files, each representing one of the four clusters that derived from the algorithm based on the establishments’ average sanitation scores.This data-driven classification of food establishments highlights opportunities for targeted inspections and improved food safety across the city.
The full report is available upon request.