[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-124378-es":3,"doc-seo-124378-110":30,"detail-sidebar-cat-0-es-110":84},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},124378,962085570644,"Evangeline","https://ap-avatar.wpscdn.com/davatar_994ba38a5ba835b3df7d355c54d3ed8d",41,"Investigación e Informes","Comparación de cinco modelos de machine learning para la predicción de las elecciones presidenciales en Colombia - una perspectiva con datos composicionales","La investigación aborda la limitada integración de técnicas de machine learning en el análisis electoral mediante un enfoque que combina cinco modelos (random forest, gradient boosting, support vector machines, k-nearest neighbors y redes neuronales feedforward) con análisis de datos composicionales. El objetivo es predecir la distribución de votos de elecciones presidenciales en Colombia a nivel municipal, representando el espectro ideológico unidimensional Izquierda-Derecha. Los modelos se entrenaron con 70% de datos de 2002 a 2022 y se evaluaron con 30% restante, mostrando desempeños comparables. Se destaca la transformación log-cociente centrada en redes neuronales feedforward, con la mayor calidad predictiva según la variabilidad observada.","Comunicaciones en Estadística  \nJulio 2025, Vol. 18, No. 1, pp. 1–27  \nComparación de cinco modelos de machine learning para lapredicción de las elecciones presidenciales en Colombia: una perspectiva con datos composicionales  \nComparison of five machine learning models for the prediction of presidential elections in  \nColombia: a compositional data perspective  \nPaula Andrea Leal Varón.a[paulalealv@usantotomas.edu.co](paulalealv@usantotomas.edu.co)  \nGermán Andrés Galeano Ortiz.b  \n[germangaleano@usantotomas.edu.co](germangaleano@usantotomas.edu.co)[ ](germangaleano@usantotomas.edu.co)Wilmer Pineda-Rios.c  \n[wilmerpineda@usta.edu.co](wilmerpineda@usta.edu.co)  \nResumen  \nEn los últimos años, numerosas investigaciones han empleado técnicas de machine learning y análisis de datos composicionales en distintos campos de estudio. Sin embargo, su integración en el análisis electoral sigue siendo escasa. Por tal razón, este trabajo integra ambos enfoques aplicando cinco modelos de machine learning: random forest, gradient boosting, support vector machines, k-nearest neighbors, y feedforward neural networks, para predecir los resultados de las elecciones presidenciales en Colombia a nivel municipal, considerando los datos como composicionales. Específicamente, se pronostica la distribuciónde votos de cada municipio en el espectro ideológico unidimensional Izquierda-Derecha. De esta forma, se busca no solo mejorar la precisión de las predicciones, sino también generar un avance importanteen las metodologías aplicadas al análisis electoral. Los modelos se entrenaron con el 70% de los datos de las elecciones presidenciales entre 2002 y 2022, y se evaluó su rendimiento en el 30% restante. Los algoritmos mostraron desempeños similares entre las transformaciones de cada espectro ideológico con porcentajes de variabilidad entre el 56% y 94% en la predicción de la proporción de votos, destacándose el modelo de feedforward neural networks con la transformación log-cociente centrada, que alcanzó los mejores resultados.  \n–  \nPalabras clave: elecciones precidenciales, machine-learning, random forest, gradient boosting, support vector machines, k-nearest neighbors, feedforward neural networks, datos composicionales, logaritmos decocientes.  \nAbstract  \nIn recent years, numerous studies have employed machine learning techniques and compositional data analysis in various fields of study. However, their integration into electoral analysis remains limited. For this reason, this work combines both approaches by applying five machine learning models: random forest, gradient boosting, support vector machines, k-nearest neighbors, and feedforward neural networks, to predict the results of the presidential elections in Colombia at the municipal level, onsidering the data as compositional. Specifically, it forecasts the vote distribution in each municipality along a unidimensional  \na Estudiante, Maestría en Estadística Aplicada, Universidad Santo Tomás, Bogotá bEstudiante, Maestría en Estadística Aplicada, Universidad Santo Tomás, Bogotá cDocente, Facultad de Estadística, Universidad Santo Tomás, Bogotá  \n2 Paula Andrea Leal Varón., Germán Andrés Galeano Ortiz. & Wilmer Pineda-Rios.  \nLeft-Right ideological spectrum. This approach aims not only to improve prediction accuracy but also to comtribute a significant advancement in methodologies applied to electoral analysis. The models were trained on 70% of the presidential election data from 2002 to 2022 and evaluated on the remaining 30% . The algorithms demonstrated similar performance across transformations of each ideological spectrum, with variability percentages between 56% and 94% in predicting vote proportions, with the feedforward  \nneural networks model using the centered log-ratio transformation achieving the best results.– Keywords: presidential elections, machine-learning, random forest, gradient boosting, support vector machines, k-nearest neighbors, feedforward neural networks, compositional ","cbCaivTNZdsLMe5I","https://ap.wps.com/l/cbCaivTNZdsLMe5I","pdf",1304751,1,27,"Spanish","es",110,"# Introducción\n## Antecedentes y motivación\n## Enfoque metodológico y modelos","[{\"question\":\"Cómo se entrenan y evalúan los modelos?\",\"answer\":\"Los modelos se entrenan con el 70% de los datos de 2002 a 2022 y se evalúa su rendimiento con el 30% restante, analizando la variabilidad en la predicción de proporciones de voto.\"}]","Comparación de cinco modelos de machine learning para la predicción de las elecciones presidenciales en Colombia - una perspectiva con datos composicionales | PDF",1785821888,42,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":79,"head_meta":81,"extra_data":83,"updated_unix":28},"comparison-of-five-machine-learning-models-for-predicting-presidential-elections-in-colombia-a-compositional-data-perspective","",{"@graph":36,"@context":78},[37,54,69],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/es/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/es/document/investigación-e-informes/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/es/document/comparison-of-five-machine-learning-models-for-predicting-presidential-elections-in-colombia-a-compositional-data-perspective/124378/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":63,"encodingFormat":61,"isAccessibleForFree":64,"interactionStatistic":65},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-09","2026-08-04",true,{"@type":66,"interactionType":67,"userInteractionCount":20},"InteractionCounter",{"@type":68},"ViewAction",{"@type":70,"mainEntity":71},"FAQPage",[72],{"name":73,"@type":74,"acceptedAnswer":75},"Cómo se entrenan y evalúan los modelos?","Question",{"text":76,"@type":77},"Los modelos se entrenan con el 70% de los datos de 2002 a 2022 y se evalúa su rendimiento con el 30% restante, analizando la variabilidad en la predicción de proporciones de voto.","Answer","https://schema.org",{"og:url":52,"og:type":80,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":82,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":85},[86,91,95,99,103,105,109,113,116,120],{"id":87,"doc_module":4,"doc_module_name":46,"category_name":88,"show_sort_weight":89,"slug":90},39,"Cómic",60,"comic",{"id":92,"doc_module":4,"doc_module_name":46,"category_name":93,"show_sort_weight":89,"slug":94},43,"Estilo de Vida","lifestyle",{"id":96,"doc_module":4,"doc_module_name":46,"category_name":97,"show_sort_weight":89,"slug":98},38,"Examen","exam",{"id":100,"doc_module":4,"doc_module_name":46,"category_name":101,"show_sort_weight":89,"slug":102},44,"General","general",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":89,"slug":104},"research-report",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":89,"slug":108},37,"Literatura","literature",{"id":110,"doc_module":4,"doc_module_name":46,"category_name":111,"show_sort_weight":89,"slug":112},22,"Relatos y Novelas","story-novel",{"id":29,"doc_module":4,"doc_module_name":46,"category_name":114,"show_sort_weight":89,"slug":115},"Religión y Espiritualidad","religion-spirituality",{"id":117,"doc_module":4,"doc_module_name":46,"category_name":118,"show_sort_weight":89,"slug":119},40,"Salud y Atención Médica","healthcare",{"id":121,"doc_module":4,"doc_module_name":46,"category_name":122,"show_sort_weight":89,"slug":123},24,"Tecnología","technology"]