CRISP-DM-BASED MACHINE LEARNING MODELS FOR ANALYZING THE DEPRESSION LEVEL IN STUDENTS OF THE NATIONAL POLYTECHNIC SCHOOL

This project analyzes the depression rates among students from Escuela Politécnica Nacional (EPN). A total of 302 students from different EPN careers, voluntarily and anonymously completed an online survey of the Beck Depression Inventory-II (BDI-II). In addition, they were asked to answer 19 questions related to the lifestyle of an EPN student; These questions were reviewed and endorsed about their possible relationship with depressive disorders by a professional in the field of psychology. The CRISP-DM methodology was used for the project phases, which involved the analysis of the current situation, objectives setting, data collection, data preparation, and construction of ML models that allows predicting the degree of depression based on the BDI-II metrics and evaluation of the models. The model obtained has 0.59 accuracy score and shows that variables of gender, age and relationships are significant to determine severity depression.