K-mean Clustering: a case study in Yvelines, Île-de-France

Cluster analysis is widely used in various fields to classify data that have similarities into the same group, but are different from the objects in the other group, to gain insights into various applications such as marketing, urban planning, fraud detection, biology, and many more. The ability to cluster the number of fire operations in the Île-de-France, especially in Yvelines, will definitely help the fire and rescue service to make better decisions in emergency response and increase the efficiency of material and human resources, which, if they can be reduced, can lower financial costs. In this paper, a collection of fire operations in different departments of the Île-de-France was made, but only the information on the 20 largest departments of Yvelines was selected. After choosing the optimal k-value, the K-means clustering method was applied, and further research was carried out to summarize the criteria by which the Insee were classified into clusters. Breakpoints were detected, statistics were obtained for each insee, linear regression was implemented, and finally time series decomposition was performed. The results show that the clusters are well separated and have a dense grouping of insee with approximately the same trend, meaning the same number of interventions.