Modeling and learning context enriched behavior patterns in ambient assisted living

"Automatic learning of personal routines can provide better insights about one's behavior and help detecting behavior changes that can signal health or well-being deterioration. In this thesis, we address the problem of representing personal routines and their variations, called behavior patterns, and learning them from observations collected by sensors in a smart home. Behavior patterns can change depending on context conditions such as day of week or weather. Not considering such variations leads to unnoticed patterns or to incorrect interpretations of them. We propose Context Enriched Behavior Patterns; behavior patterns that are described with the context they are valid on. The contributions of this thesis include: (1) a formal model to represent such patterns, (2) a learning algorithm that automatically detects behavior patterns and their variations in a single data scan and incrementally, (3) a publicly available dataset for evaluating this and other smart home proposals, (4) the evaluation of the proposal in real-life datasets."--Tomado del Formato de Documento de Grado

• Autores: Lago Martínez, Paula Andrea
• Tipo de recurso: Doctoral thesis
• Fecha de publicación: 2017
• Institución: Universidad de los Andes
• Repositorio: Séneca: repositorio Uniandes
• Idioma: eng
• OAI Identifier: oai:repositorio.uniandes.edu.co:1992/38675
• Acceso en línea: http://hdl.handle.net/1992/38675
• Palabra clave: Aprendizaje automático (Inteligencia artificial) - Investigaciones - Estudio de casos; Dispositivos de autoayuda para ancianos - Innovaciones tecnológicas - Investigaciones; Tecnología de asistencia en cómputo - Investigaciones; Algoritmos evolutivos; Ingeniería
• Rights: openAccess
• License: https://repositorio.uniandes.edu.co/static/pdf/aceptacion_uso_es.pdf