Adaptive model to predict sleep quality from the personalized selection of sleep hygiene factors : an approach from the machine learning algorithms [recurso electrónico] / Arturo Jesús Laflor Hernández ; dirigida por Mabel Vázquez Briseño
Tipo de material:
TextoDetalles de publicación: Ensenada, Baja California, 2018Descripción: 1 recurso en línea xvi, 129 p. : ilTema(s): Ingeniería -- Tesis y disertaciones académicas | Algoritmos (computacionales)Clasificación LoC:QA76.9 .A43 | L33 2018Recursos en línea: Tesis Digital
Nota de disertación: Tesis (Doctorado)--Universidad Autónoma de Baja California. Facultad de Ingeniería, Arquitectura y Diseño, Ensenada, 2018 Resumen: Many social and behavioral phenomena difficult to study years ago, today find explanation when they model them through Machine Learning Algorithms. This is possible by the daily interaction between people and devises capturing and storing personal and environmental data. Capacity of memory, process and ability to interact with remote and powerful servers performing the heavy work enable the analysis of large amounts of data with velocity and precision. Phenomena referred to above depends on many features and their relations which they sometimes
are not linear. We presented an adaptive model to predict the Sleep Quality from the critical Sleep Hygiene Factors (SHF) selection. Methodology involved a transactional study to identify the critical SHF in the study population, and a longitudinal study to validate the generalization of the selected SHF to particular cases. The no generalization of the SHF forces us to consider the complete set of the SHF as source data and to design into the predictive model a feature selection
algorithm (fsXLR) to discriminate the non-relevant SHF. fsXLR extracts the critical factors finding most variance explanation through the implementation of the Bagging and Best-Search techniques to the results of the three well known algorithms XGBOOST, LASSO and RF.
| Tipo de ítem | Biblioteca actual | Colección | Signatura | Copia número | Estado | Fecha de vencimiento | Código de barras |
|---|---|---|---|---|---|---|---|
| Tesis | Biblioteca Central Ensenada | Colección de Tesis | QA76.9 .A43 L33 2018 (Browse shelf(Abre debajo)) | 1 | Disponible | ENS089055 |
Navegando Biblioteca Central Ensenada Estantes, Código de colección: Colección de Tesis Cerrar el navegador de estanterías (Oculta el navegador de estanterías)
Tesis (Doctorado)--Universidad Autónoma de Baja California. Facultad de Ingeniería, Arquitectura y Diseño, Ensenada, 2018
Incluye referencias bibliográficas
Many social and behavioral phenomena difficult to study years ago, today find explanation when they model them through Machine Learning Algorithms. This is possible by the daily interaction between people and devises capturing and storing personal and environmental data. Capacity of memory, process and ability to interact with remote and powerful servers performing the heavy work enable the analysis of large amounts of data with velocity and precision. Phenomena referred to above depends on many features and their relations which they sometimes
are not linear. We presented an adaptive model to predict the Sleep Quality from the critical Sleep Hygiene Factors (SHF) selection. Methodology involved a transactional study to identify the critical SHF in the study population, and a longitudinal study to validate the generalization of the selected SHF to particular cases. The no generalization of the SHF forces us to consider the complete set of the SHF as source data and to design into the predictive model a feature selection
algorithm (fsXLR) to discriminate the non-relevant SHF. fsXLR extracts the critical factors finding most variance explanation through the implementation of the Bagging and Best-Search techniques to the results of the three well known algorithms XGBOOST, LASSO and RF.
