Ajuste de modelos lineales mixtos para la estimación futura del consumo de energía en servicios públicos de clientes del departamento de Antioquia

dc.contributor.advisorHernández Barajas, Freddy
dc.contributor.authorCandela Aristizábal , Daniel Alberto
dc.contributor.orcidHernández-Barajas, Freddy [0000000174593329]
dc.coverage.regionAntioquia (Colombia(
dc.date.accessioned2026-03-13T19:31:57Z
dc.date.available2026-03-13T19:31:57Z
dc.date.issued2026-02-26
dc.descriptionIlustraciones
dc.description.abstractEn un mundo cada vez más orientado hacia el análisis de datos, las herramientas estadísticas brindan una ventaja competitiva significativa al permitir un enfoque informado y basado en evidencia para la toma de decisiones. Con este enfoque, las empresas pueden tomar decisiones más informadas y desarrollar estrategias sólidas para abordar los desafíos y oportunidades emergentes. Las metodologías estadísticas pueden aplicarse en una amplia variedad de campos y sectores empresariales, lo que les brinda una versatilidad única. En el caso particular de las empresas prestadoras de servicios públicos domiciliarios en Colombia, la normatividad exige la revisión de las variaciones en el consumo de sus clientes. Sin embargo, en la identificación de estas variaciones, conocidas como desviaciones significativas, se les permite a las empresas aplicar su criterio. La mayoría de ellas utiliza cálculos de variación del consumo del mes actual frente al consumo promedio de los últimos seis meses. Aunque esta práctica es común, son pocos los casos en los que se han propuesto metodologías basadas en comportamientos históricos respaldados por modelos estadísticos solidos. Por lo tanto, en este contexto, se propone un enfoque que busca mejorar la precisión de las predicciones del consumo de cada cliente. Se propone un modelo predictivo basado en el ajuste de Modelos Lineales Mixtos (MLM). Los MLM son una herramienta estadística que permite tomar en cuenta efectos fijos y aleatorios, lo que puede mejorar significativamente la precisión de las predicciones en comparación con enfoques más simples. Para esto, se tomó un histórico de consumos de energía eléctrica de 74,421 clientes del departamento de Antioquia, desde junio de 2024 hasta julio de 2025, y se ajustaron y evaluaron modelos basados en MLM para la identificación de desviaciones significativas. Adicionalmente, se creó un dashboard en R para permitir que los clientes internos de la empresa puedan utilizar el modelo estadístico ajustado para identificar consumos de energía. (Texto tomado de la fuente)spa
dc.description.abstractIn an increasingly data-driven world, statistical tools provide a significant competitive advantage by enabling an informed and evidence-based approach to decision-making. With this approach, companies can make more informed decisions and develop robust strategies to address emerging challenges and opportunities. Statistical methodologies can be applied across a wide range of fields and business sectors, providing them with unique versatility. In the specific case of utility service companies in Colombia, regulations mandate the review of variations in customer consumption. However, in dentifying these variations, known as significant deviations, companies are allowed to apply their judgment. Most of them use calculations of the current month’s consumption versus the average consumption of the last six months. Although this practice is common, few cases propose methodologies based on historical behaviors supported by strong statistical models. Therefore, in this context, an approach is proposed to enhance the accuracy of predictions for each customer’s consumption. A predictive model based on the adjustment of Mixed Linear Models (MLM) is suggested. MLMs are a statistical tool that allows for the consideration of both fixed and random effects, which can significantly improve prediction accuracy compared to simpler approaches. For this purpose, a history of electricity consumption from 522 customers in the Antioquia department from June 2024 to July 2025 was collected, and models based on MLM was fitted and evaluated for the identification of significant deviations. Additionally, an R-based dashboard was developed to enable the company’s internal clients to utilize the fitted statistical model for identifying energy consumption patterns.eng
dc.description.curricularareaEstadística.Sede Medellín
dc.description.degreelevelMaestría
dc.description.degreenameMagíster en Ciencias - Estadística
dc.description.researchareaEstadística aplicada
dc.format.extent1 recurso en línea (116 páginas)
dc.format.mimetypeapplication/pdf
dc.identifier.instnameUniversidad Nacional de Colombiaspa
dc.identifier.reponameRepositorio Institucional Universidad Nacional de Colombiaspa
dc.identifier.repourlhttps://repositorio.unal.edu.co/spa
dc.identifier.urihttps://repositorio.unal.edu.co/handle/unal/89754
dc.language.isospa
dc.publisherUniversidad Nacional de Colombia
dc.publisher.branchUniversidad Nacional de Colombia - Sede Medellín
dc.publisher.facultyFacultad de Ciencias
dc.publisher.placeMedellín, Colombia
dc.publisher.programMedellín - Ciencias - Maestría en Ciencias - Estadística
dc.relation.referencesAkaike, H. (1974). A new look at the statistical model identification. IEEE Transactions on Automatic Control , 19 (6), 716–723.
dc.relation.referencesAl-Hamoud, L. A. (2016). A review of techniques and algorithms for electric load forecasting: A comparative analysis. International Journal of Electrical Power Energy Systems, 82 , 319–336.
dc.relation.referencesAparicio, J., Garc´ıa-Mar´ı, R., y Jim´enez-Barrera, S. (2012). Modeling residential electricity consumption with mixed-effects models. Energy and Buildings, 49 , 477–485.
dc.relation.referencesArellano, M. (2003). Panel data econometrics. Oxford, UK: Oxford University Press.
dc.relation.referencesBaayen, R. H. (2008). Mixed-effects models in linguistics with r. New York, NY: Springer.
dc.relation.referencesBall, R. (2022). lmesplines: Add smoothing spline modelling capability to ‘nlme’ [Manual de software inform´atico]. Descargado de https://CRAN.R-project.org/package= lmeSplines (R package version 1.1-12)
dc.relation.referencesBates, D., M¨achler, M., Bolker, B., y Walker, S. (2015). Fitting linear mixed-effects models using lme4. Journal of Statistical Software, 67 (1), 1–48. doi: 10.18637/jss.v067.i01
dc.relation.referencesBox, G. E. P., Jenkins, G. M., y Reinsel, G. C. (1994). Time series analysis: Forecasting and control (3rd ed.). Englewood Cliffs, NJ: Prentice Hall.
dc.relation.referencesBrooks, M. E., Kristensen, K., van Benthem, K. J., Magnusson, A., Berg, C. W., Nielsen, A., . . . Bolker, B. M. (2017). glmmTMB balances speed and flexibility among packages for zero-inflated generalized linear mixed modeling. The R Journal, 9 (2), 378–400. doi: 10.32614/RJ-2017-066
dc.relation.referencesBurkner, P.-C. (2021). Bayesian item response modeling in R with brms and Stan. Journal of Statistical Software, 100 (5), 1–54. doi: 10.18637/jss.v100.i05
dc.relation.referencesCasella, G., y Berger, R. L. (2021). Statistical inference. Cengage Learning.
dc.relation.referencesChang, W., Cheng, J., Allaire, J., Sievert, C., Schloerke, B., Xie, Y., . . . de Vries, A. (2024). shiny: Web application framework for r [Manual de software inform´atico]. Descargado de https://CRAN.R-project.org/package=shiny (R package version 1.9.1)
dc.relation.referencesCharytoniuk, W., Chen, M., y Van Olst, F. (1998). Neural network approach to short-term load forecasting. IEEE Transactions on Power Systems, 13 (3), 938–944.
dc.relation.referencesChatfield, C. (2000). Time-series forecasting (1st ed.). Boca Raton: Chapman and Hall/ CRC.
dc.relation.referencesDeb, C., Zhang, F., Yang, J., Lee, S. E., y Shah, N. (2017). A review of energy forecasting models for smart grid development. Renewable and Sustainable Energy Reviews, 72 , 1129–1143.
dc.relation.referencesFan, J., Zhao, J., y Li, X. (2018). Mixed-effects models for understanding household electricity load profiles from smart meter data. Applied Energy, 211 , 1468–1479.
dc.relation.referencesFisher, R. A. (1925). Statistical methods for research workers. Oliver and Boyd.
dc.relation.referencesFitzmaurice, G. M., Laird, N. M., y Ware, J. H. (2011). Applied longitudinal data analysis: Modeling change and event occurrence (2nd ed.). Hoboken, NJ: Wiley-Blackwell.
dc.relation.referencesGardner Jr, E. S. (1985). Exponential smoothing: The state of the art. Journal of Forecasting, 4 (1), 1–28.
dc.relation.referencesGeller, N. L. (2006). Multilevel modeling of health data. Statistical Methods in Medical Research, 15 (3), 201–223.
dc.relation.referencesGelman, A., Carlin, J. B., Stern, H. S., Dunson, D. B., Vehtari, A., y Rubin, D. B. (2013). Bayesian data analysis. CRC press.
dc.relation.referencesGelman, A., y Hill, J. (2006). Data analysis using regression and multilevel/hierarchical models. New York, NY: Cambridge University Press.
dc.relation.referencesHadfield, J. D. (2010). Mcmc methods for multi-response generalized linear mixed models: The MCMCglmm R package. Journal of Statistical Software, 33 (2), 1–22. Descargado de https://www.jstatsoft.org/v33/i02/
dc.relation.referencesHartig, F. (2022). DHARMa: Residual diagnostics for hierarchical (multi-level / mixed) regression models [Manual de software inform´atico]. Descargado de https://cran.r -project.org/package=DHARMa (R package version 0.4.6)
dc.relation.referencesHarville, D. A. (1977). Maximum likelihood approaches to variance component estimation and to related problems. Journal of the American Statistical Association, 72 (358), 320–338.
dc.relation.referencesHastie, T., Tibshirani, R., y Friedman, J. H. (2009). The elements of statistical learning: data mining, inference, and prediction (Vol. 2). Springer.
dc.relation.referencesHernández, F. (2025). Modelos mixtos con R [Manual de software informático]. Medellín, Colombia. Descargado de https://fhernanb.github.io/libro modelos mixtos/IC .html (Consultado el 20 de febrero de 2026)
dc.relation.referencesHoffman, L. (2007). Longitudinal data analysis. Hoboken, NJ: Wiley-Blackwell.
dc.relation.referencesHong, Y. (2020). Forecasting for economics and business. Singapore: Springer Nature.
dc.relation.referencesHyndman, R. J., y Athanasopoulos, G. (2018). Forecasting: Principles and practice. OTexts. Descargado de https://otexts.com/fpp2/
dc.relation.referencesHyndman, R. J., Koehler, A. B., Ord, J. K., y Snyder, R. D. (2008). Forecasting with exponential smoothing: The state space approach. Berlin, Heidelberg: Springer Science & Business Media.
dc.relation.referencesJames, G., Witten, D., Hastie, T., y Tibshirani, R. (2013). An introduction to statistical learning: with applications in r. New York, NY: Springer.
dc.relation.referencesKim, D.-H., Lee, J.-Y., y Kim, B.-S. (2016). Applying mixed-effects models to industrial energy demand forecasting. Energy Economics, 54 , 220–231.
dc.relation.referencesKuznetsova, A., Brockhoff, P. B., y Christensen, R. H. B. (2017). lmerTest package: Tests in linear mixed effects models. Journal of Statistical Software, 82 (13), 1–26. doi: 10.18637/jss.v082.i13
dc.relation.referencesLaird, N. M., y Ware, J. H. (1982a). Random-effects models for longitudinal data. Biometrics, 38 (4), 963–974. Descargado de http://www.jstor.org/stable/2529876
dc.relation.referencesLaird, N. M., y Ware, J. H. (1982b). Random-effects models for longitudinal data. Biometrics, 963–974.
dc.relation.referencesLefcheck, J. S. (2021). piecewisesem: Piecewise structural equation modeling [Manual de software inform´atico]. Descargado de https://cran.r-project.org/package= piecewiseSEM (R package version 2.2)
dc.relation.referencesLi, J., Zhang, W., y Ding, M. (2019). Short-term load forecasting using spatiotemporal mixed effects model. Electric Power Systems Research, 176 , 105953.
dc.relation.referencesLong, J. S. (1997). Regression models for categorical and limited dependent variables. Thousand Oaks, CA: SAGE Publications.
dc.relation.referencesLuke, D. A. (2004). Multilevel modeling. Thousand Oaks, CA: SAGE Publications.
dc.relation.referencesMcCulloch, C. E., Searle, S. R., y Neuhaus, J. M. (2008). Generalized, linear, and mixed models (2nd ed.). Hoboken, NJ: John Wiley & Sons.
dc.relation.referencesMolenberghs, G., Verbeke, G., y cols. (2005). Models for discrete longitudinal data. New York: Springer.
dc.relation.referencesMontgomery, D. C., Peck, E. A., y Vining, G. G. (2012). Applied linear statistical models. John Wiley & Sons.
dc.relation.referencesNakagawa, S., Schielzeth, H., y O’Hara, R. B. (2017). R2 for mixed models. Methods in Ecology and Evolution, 8 (10), 1170–1178.
dc.relation.referencesNieuwenhuis, R., Te Grotenhuis, M., y Pelzer, B. (2012). influence.me: Tools for detecting influential data in mixed effects models. R Journal, 4 (2), 38-47.
dc.relation.referencesPerry, P. O. (2017). Fast moment-based estimation for hierarchical models. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 79 , 267-291. doi: 10.1111/rssb.12165
dc.relation.referencesPinheiro, J., Bates, D., y R Core Team. (2023). nlme: Linear and nonlinear mixed effects models [Manual de software inform´atico]. Descargado de https://CRAN.R-project .org/package=nlme (R package version 3.1-162)
dc.relation.referencesPinheiro, J. C., y Bates, D. M. (2000). Mixed-effects models in s and s-plus. New York, NY: Springer Science & Business Media.
dc.relation.referencesPullinger, D., y Bell, M. (2013). Mixed models for predicting electricity demand in commercial buildings. Energy and Buildings, 63 , 87–94.
dc.relation.referencesRaudenbush, S. W., y Bryk, A. S. (2002). Hierarchical linear models: Applications and data analysis methods (2nd ed.). Thousand Oaks, CA: SAGE Publications.
dc.relation.referencesRosseel, Y. (2012). lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48 (2), 1–36. doi: 10.18637/jss.v048.i02
dc.relation.referencesSafken, B., Rugamer, D., Kneib, T., y Greven, S. (2021). Conditional model selection in mixed-effects models with cAIC4. Journal of Statistical Software, 99 (8), 1–30. doi: 10.18637/jss.v099.i08
dc.relation.referencesSchwarz, G. (1978). Estimating the dimension of a model. The Annals of Statistics, 6 (2), 461–464. doi: 10.1214/aos/1176344136
dc.relation.referencesShumway, R. H., y Stoffer, D. S. (2017). Time series analysis and its applications: With r examples (4th ed.). New York, NY: Springer.
dc.relation.referencesSinger, J. D., y Willett, J. B. (2003). Applied longitudinal data analysis: Modeling change and event occurrence. New York, NY: Oxford University Press.
dc.relation.referencesSnijders, T. A. B., y Bosker, R. J. (2011). Multilevel analysis: An introduction to basic and advanced multilevel modeling (2nd ed.). London: SAGE Publications.
dc.relation.referencesStapper-Buitrago, C. (2009). La regulación de los servicios públicos domiciliarios: eficiencia, equidad, y protección de los derechos de los usuarios. Comisión de Regulación de Agua Potable y Saneamiento Básico-CRA.
dc.relation.referencesStroup, W. W. (2012). Generalized linear mixed models: modern concepts, methods and applications. CRC press.
dc.relation.referencesTaylor, S. J. (2003). Financial modelling with time series analysis: an introduction to arima modeling. The Journal of Business, 76 (1), 1–20.
dc.relation.referencesVenables, W. N., y Ripley, B. D. (2021). MASS: Support functions and datasets for venables and ripley’s MASS [Manual de software inform´atico]. Descargado de https://cran.r -project.org/package=MASS (R package version 7.3-53.1)
dc.relation.referencesWang, Y., Cai, W., y Peng, J. (2014). Exploring residential energy consumption patterns in urban areas using mixed-effects models. Journal of Cleaner Production, 82 , 15–24.
dc.relation.referencesWooldridge, J. M. (2016). Introductory econometrics: A modern approach (6th ed.). Boston, MA: Cengage Learning.
dc.relation.referencesZhao, R., Qian, B., y Zhang, L. (2017). A hybrid short-term load forecasting model using mixed-effect trees. Energy, 134 , 490–500.
dc.relation.referencesZheng, J., Wang, Y., y Li, G. (2017). Mixed effects model for hourly electricity consumption forecasting of households. Energy and Buildings, 150 , 332–342.
dc.rights.accessrightsinfo:eu-repo/semantics/openAccess
dc.rights.licenseAtribución-NoComercial-CompartirIgual 4.0 Internacional
dc.rights.urihttp://creativecommons.org/licenses/by-nc-sa/4.0/
dc.subject.ddc510 - Matemáticas::519 - Probabilidades y matemáticas aplicadas
dc.subject.ddc330 - Economía::333 - Economía de la tierra y de la energía
dc.subject.lembServicios públicos domiciliarios
dc.subject.lembModelos lineales (Estadística)
dc.subject.lembProbabilidades
dc.subject.lembConsumo de energía - Antioquia (colombia)
dc.subject.lembDemanda de energía - Antioquia (Colombia)
dc.subject.proposalSignificant deviationspa
dc.subject.proposalSignificant deviationspa
dc.subject.proposalSignificant deviationeng
dc.subject.proposalSignificant deviationeng
dc.subject.proposalSignificant deviationeng
dc.subject.proposalSignificant deviationeng
dc.subject.proposalSignificant deviationeng
dc.titleAjuste de modelos lineales mixtos para la estimación futura del consumo de energía en servicios públicos de clientes del departamento de Antioquiaspa
dc.title.translatedAdjustment of mixed linear models for the future estimation of energy consumption in public utilities for customers in the department of Antioquia.eng
dc.typeTrabajo de grado - Maestría
dc.type.coarhttp://purl.org/coar/resource_type/c_bdcc
dc.type.coarversionhttp://purl.org/coar/version/c_ab4af688f83e57aa
dc.type.contentText
dc.type.driverinfo:eu-repo/semantics/masterThesis
dc.type.redcolhttp://purl.org/redcol/resource_type/TM
dc.type.versioninfo:eu-repo/semantics/acceptedVersion
dcterms.audience.professionaldevelopmentEstudiantes
dcterms.audience.professionaldevelopmentInvestigadores
oaire.accessrightshttp://purl.org/coar/access_right/c_abf2

Archivos

Bloque original

Mostrando 1 - 1 de 1
Cargando...
Miniatura
Nombre:
1017178988.2026.pdf
Tamaño:
16.13 MB
Formato:
Adobe Portable Document Format
Descripción:
Tesis de Maestría en Ciencias - Estadística

Bloque de licencias

Mostrando 1 - 1 de 1
Cargando...
Miniatura
Nombre:
license.txt
Tamaño:
5.74 KB
Formato:
Item-specific license agreed upon to submission
Descripción: