Modelo predictivo para el reconocimiento de actividades humanas basado en técnicas de Machine Learning y de selección de características

Ambient assisted living (AAL), focus on generating innovative products and services in order to aid and medical attention to elderly people who suffer from neurodegenerative diseases or a disability. This research area is responsible for the development of activity recognition systems (ARS) which are based on Human Activity Recognition (HAR), specifically in activities of daily life (ADL) in indoor environments. These systems make it possible to identify the type of activity that people carry out, offering a possibility of effective assistance that allows them to carry out daily activities with total normality. The performance of the ARS in the HAR process must be evaluated through the approach of experimental scenarios with data sets available by the scientific community in online repositories, this work proposes a variety of combinations of machine learning algorithms with feature selection algorithms, obtaining as a result a functional model for the HAR, which combines the classification algorithm Logistic model trees (LMT) and the feature selection algorithm One R.

• Autores: Patiño Saucedo, Janns Álvaro
• Fecha de publicación: 2019
• Institución: Corporación Universidad de la Costa
• Repositorio: REDICUC - Repositorio CUC
• Idioma: spa
• OAI Identifier: oai:repositorio.cuc.edu.co:11323/8249
• Acceso en línea: https://hdl.handle.net/11323/8249; https://repositorio.cuc.edu.co/
• Palabra clave: Human Activity Recognition (HAR); Machine learning; Classification; Feature selection; Reconocimiento de Actividades Humanas (HAR); Aprendizaje automático; Clasificación; Selección de características
• Rights: openAccess
• License: Attribution-NonCommercial-ShareAlike 4.0 International