The issue of comfort in school classrooms in Mexico has not been adequately addressed nor effectively managed with the available tools. This inadequacy stems from a lack of proper understanding regarding their implementation. This study conducts an analysis of the application of data mining techniques utilizing climatological variables within a classroom setting to identify patterns and behaviors that predict student comfort. A comparison of various data mining models is performed, and the most effective models are presented. The study employs a dataset encompassing variables such as Temperature, Humidity, Air Quality, Light, and Noise from a classroom environment. Patterns were identified and subsequently labeled into two clusters to develop a student comfort model. Two unsupervised learning models, hierarchical clustering and k-means, were utilized alongside supervised learning models including Neural Networks, Naive Bayes, Decision Trees, and K-Nearest Neighbors.