Longitudinal analyses of childhood obesity factors

Identifier:  TFM:1894

Handle: https://hdl.handle.net/20.500.11797/TFM1894

Authors:  Práger, Zsófia

Abstract:
This master thesis explores factors contributing to childhood obesity using data from hospitals in Catalonia, focusing on a cohort of pregnant women and their children up to age 5. Key parental and infant variables were analyzed to develop a predictive model for childhood obesity. Two new experimental methods were introduced: data transformation into increments and sequential pattern representation. A total of 150 patterns were identified and analyzed using Logistic Regression, Decision Tree, and Random Forest algorithms. The findings showed that these new methods were less effective compared to other state-of-the-art approaches.