Modelos de aprendizaje profundo para la predicción de comportamiento sedentario futuro

It is well known that sedentary behavior has negative consequences for health. Therefore, encouraging individuals to avoid this type of behavior can help to reduce different risk indicators. In this work, different deep learning architectures were evaluated to predict the future sedentary behavior of an individual from the captured records of different sensors available on mobile devices. Users with different levels of energy expenditure were analyzed, and encouraging results were obtained that demonstrate the efficiency of the proposed architectures.