Browsing by Subject "Time series"

Now showing 1 - 6 of 6
  • Thumbnail Image
    Publication
    Open Access
    A methodology for energy multivariate time series forecasting in smart buildings based on feature selection
    (Elsevier, 2019-05-10) González Vidal, Aurora; Jiménez Barrionuevo, Fernando; Skarmeta Gómez, Antonio; Ingeniería de la Información y las Comunicaciones; Facultades de la UMU::Facultad de Informática
    The massive collection of data via emerging technologies like the Internet of Things (IoT) requires finding optimal ways to reduce the created features that have a potential impact on the information that can be extracted through the machine learning process. The mining of knowledge related to a concept is done on the basis of the features of data. The process of finding the best combination of features is called feature selection. In this paper we deal with multivariate time-dependent series of data points for energy forecasting in smart buildings. We propose a methodology to transform the time-dependent database into a structure that standard machine learning algorithms can process, and then, apply different types of feature selection methods for regression tasks. We used Weka for the tasks of database transformation, feature selection, regression, statistical test and forecasting. The proposed methodology improves MAE by 59.97% and RMSE by 40.75%, evaluated on training data, and it improves MAE by 42.28% and RMSE by 36.62%, evaluated on test data, on average for 1-step-ahead, 2-step-ahead and 3-step-ahead when compared to not applying any feature selection methodology.
  • Thumbnail Image
    Publication
    Open Access
    Feature selection for blood glucose level prediction in type 1 Diabetes Mellitus by using the Sequential Input Selection Algorithm (SISAL)
    (MDPI, 2019-09-14) Rodríguez Rodríguez, Ignacio ; Rodríguez, José Víctor; González Vidal, Aurora; Zamora Izquierdo, Miguel Ángel; Ingeniería de la Información y las Comunicaciones; Facultades de la UMU::Facultad de Informática
    Feature selection is a primary exercise to tackle any forecasting task. Machine learning algorithms used to predict any variable can improve their performance by lessening their computational effort with a proper dataset. Anticipating future glycemia in type 1 diabetes mellitus (DM1) patients provides a baseline in its management, and in this task, we need to carefully select data, especially now, when novel wearable devices offer more and more information. In this paper, a complete characterization of 25 diabetic people has been carried out, registering innovative variables like sleep, schedule, or heart rate in addition to other well-known ones like insulin, meal, and exercise. With this ground-breaking data compilation, we present a study of these features using the Sequential Input Selection Algorithm (SISAL), which is specially prepared for time series data. The results rank features according to their importance, regarding their relevance in blood glucose level prediction as well as indicating the most influential past values to be taken into account and distinguishing features with person-dependent behavior from others with a common performance in any patient. These ideas can be used as strategies to select data for predicting glycemia depending on the availability of computational power, required speed, or required accuracy. In conclusion, this paper tries to analyze if there exists symmetry among the different features that can affect blood glucose levels, that is, if their behavior is symmetric in terms of influence in glycemia.
  • Thumbnail Image
    Publication
    Embargo
    Integral Analysis of Circadian Rhythms
    (Yolanda Larriba, 2023-08-10) Vicente-Martinez, J; Almaida Pagán, Pedro Francisco; Martinez-Nicolas, A; Madrid, Juan A; Rol, M A; Bonmatí Carrión, María de los Ángeles; Fisiología; Centro de Investigación Biomédica en Red Fragilidad y Envejecimiento Saludable, Instituto de Salud Carlos III, Madrid, Spain
    Our biological processes do not present a static and steady picture. Instead, most of these processes oscillate throughout the 24-h day. The system that controls this circadian rhythmicity is composed of a central pacemaker and peripheral oscillators that rely on inputs and produce outputs that can be measured through appropriate techniques. The correct measurement and analysis of the func-tioning of each part of the circadian system are becoming essential in many fields of biomedicine. In this chapter, we will cover the entire process, from the acquisition of circadian data to their analyses through parametric and non-parametric methods, including the dynamic modeling of different circadian processes.
  • Thumbnail Image
    Publication
    Open Access
    Monitoring of the Implementation of a Breastfeeding Guideline for 6 Years: A Mixed-Methods Study Using an Interrupted Time Series Approach
    (2021-05) Ruzafa Martinez, Maria; Ramos Morcillo, Antonio Jesús; Harillo Acevedo, Francisco David; Enfermería
    Background: Current literature provides poor information about the implementation of health-promoting clinical practice guidelines (CPGs) and their longitudinal monitoring. Purpose: The aim of this study was to evaluate the longitudinal impact of a CPG implementation program that promotes breastfeeding, its associated quantitative and qualitative indicators, and direct costs. Design: A mixed-methods design with a longitudinal approach was utilized, with an interrupted time series design and the analysis of reports from the implementation program as the qualitative approach. Methods: The study setting was maternity and pediatric units of a health area in the Spanish health system. The implementation of a CPG for the promotion of breastfeeding was evaluated, which included a pre-implementation year (2011), 3 years of implementation (2012-2014), and 2 years of post-implementation (2015-2016). The sample was composed of mother-infant dyads. A segmented logistic regression analysis was utilized to evaluate the changes in the most important breastfeeding indicators. A deductive thematic content analysis was performed starting with quality indicators and a descriptive economic analysis. Findings: In the 6 years of monitoring, 7,842 mother-infant dyads were recorded. The results of the quantitative indicators showed the presence of four stages: baseline, gain, adjustment, and sustainability or saturation. The breast milk at the first feeding had an increasing slope in the gain stage (24% per quarter; odds ratio [OR] = 1.24, 95% confidence interval [CI] 1.12-1.37). The exclusive breastfeeding at hospital discharge showed significant changes in the period of gain (OR = 2.45, 95% CI 1.95-3.08), which was maintained in the adjustment period, with an increase of 18% in the slope of the gain stage (OR = 1.18, 95% CI 1.06-1.32). The longitudinal distribution of the qualitative indicators showed a greater concentration of indicators towards the first half of each phase. The total cost was 209,575€ ($248,670.17). Conclusions: The implementation of the breastfeeding CPG showed early, positive, and sustained results in the exclusive breastfeeding rates. The implementation implied the application of a complex intervention, with its qualitative indicators showing a wave-shaped dynamic. Clinical relevance: Our findings contribute to the understanding and evolution of the main indicators of the implementation of a breastfeeding CPG, providing details on the magnitude of the effect, the process of change, and the associated costs.
  • Thumbnail Image
    Publication
    Open Access
    Revisión y prospectiva de la producción española en tesis doctorales de pedagogía (1976-2006)
    (Universidad de Murcia. Servicio de Publicaciones, 2008-01-01) Fernández Cano, Antonio; Torralbo Rodríguez, Manuel; Vallejo Ruiz, Mónica
    This paper reviews and analyses the last thirty years in terms of the annual production of PhD theses in the field of Pedagogy, wich are included in the Spanish database TESEO. Using deterministic models and ARIMA methodology, this study aims to define patterns of growth, verify data fit to scientometric standards and undertake prospective analysis in order to determine the trends that steer research into Pedagogy. Furthermore, qualitative considerations are provided, based on the status and production of PhD theses in Spain, wich might help to understand and improve this area of university education, at light of its integration onto the European Space for Higher Education.
  • Thumbnail Image
    Publication
    Open Access
    The dynamics of the university impact on YouTube: a comparative analysis
    (2021) Ros Gálvez, Alejandro; Meseguer Martínez, Ángel; López Buenache, Germán; Métodos Cuantitativos para la Economía y la Empresa
    The impact of universities on social networks has been widely studied, especially on Facebook and Twitter. However, there is a clear lack of research on YouTube, despite an overwhelming presence of universities on this online video platform. The objective of this work is to analyse similarities in the dynamics of views and likes between the YouTube channel of a university, Polytechnic University of Valencia (UPV), and an educational channel, the saurabhschool [25]. This is the first work that analyzes the dynamics of online video impact of a university on YouTube, which are subsequently compared to those of an educational channel to find common patterns. The times series of views and likes are obtained for both channels, their seasonal components are calculated and compared. Observation is subsequently supported by an analysis of correlations and the Euclidean distance. Results suggest that the video impact dynamics of a university channel behave similarly to those of an educational channel. These results can help universities anticipate the behavior pattern of their videos in order to maximize the impact of their content through YouTube.