Análisis comparativo de 3 algoritmos de predicción que permita analizar coordenadas georreferenciales de objetos en movimiento.
Fecha
2021
Título de la revista
ISSN de la revista
Título del volumen
Editor
Universidad de Guayaquil. Facultad de Ciencias Matemáticas y Físicas. Carrera de Ingeniería en Sistemas Computacionales.
Resumen
El trabajo de titulación se basa en realizar un análisis comparativo de 3 algoritmo de predicción (Regresión lineal, Redes neuronales y KNN), que permitan estudiar informaciones de coordenadas georreferenciales de objetos en movimiento, ya que actualmente existen problemas entorno al congestionamiento vehicular cerca de la Universidad de Guayaquil, causando que sus estudiantes y personal administrativo se retrasen en cumplir con sus obligaciones por este tipo de inconveniente. Debido a esto, se estableció como objetivo principal implementar un algoritmo que cuente con mayor exactitud y efectividad de predicción en un módulo que se alojará en el SISTEMA INTELIGENTE PARA LA MOVILIDAD SOSTENIBLE DE LA UNIVERSIDAD DE GUAYAQUIL(SIAMS-UG), haciendo uso de herramientas open Source que permitan trabajar con Machine Learning para así lograr analizar los pronósticos de los congestionamientos que se forman en los alrededores de la Universidad de Guayaquil. Para llevar a cabo el proyecto se realizó una investigación exhaustiva mediante la metodología de recolección de citas bibliográficas por medio de una revisión sistemática acerca del área de estudio definido, por consiguiente, para el procesamiento de la información se utilizaron paqueterías de datos de tipo cuantitativo para obtener las probabilidades necesarias. En cuanto al desarrollo, se empleó la metodología cascada ya que al ser un modelo lineal y de simple implementación logró enfatizar cada fase del proyecto permitiendo tener como resultados luego de varios análisis comparativos que el algoritmo más eficiente fuera el de redes neuronales, siendo así se procedió a la implementación del mismo en la aplicación web SIAMS-UG con el fin de cumplir con el objetivo planteado del problema que es obtener mejoras en la movilidad sostenible.
The degree work is based on a comparative analysis of 3 prediction algorithms (Linear Regression, Neural Networks and KNN), which allow to study information of georeferential coordinates of moving objects, since currently there are problems around the traffic congestion near the University of Guayaquil, causing that students and administrative staff are delayed in fulfilling their obligations by this type of inconvenience. Due to this, it was established as main objective to implement an algorithm that has greater accuracy and predictive effectiveness in a module that will be hosted in the INTELLIGENT SYSTEM FOR SUSTAINABLE MOBILITY OF THE UNIVERSITY OF GUAYAQUIL (SIAMS-UG), making use of open Source tools that allow working with Machine Learning to analyze the forecasts of congestion that form around the University of Guayaquil. In order to carry out the project, an exhaustive research was carried out through the methodology of collecting bibliographic citations by means of a systematic review about the defined study area, consequently, for the processing of the information, quantitative data packages were used to obtain the necessary probabilities. As for the development, the cascade methodology was used since being a linear and simple implementation model, it was able to emphasize each phase of the project allowing to have as results after several comparative analyses that the most efficient algorithm was the neural network algorithm, thus proceeding to its implementation in the SIAMS-UG web application in order to meet the stated objective of the problem, which is to obtain improvements in sustainable mobility.
The degree work is based on a comparative analysis of 3 prediction algorithms (Linear Regression, Neural Networks and KNN), which allow to study information of georeferential coordinates of moving objects, since currently there are problems around the traffic congestion near the University of Guayaquil, causing that students and administrative staff are delayed in fulfilling their obligations by this type of inconvenience. Due to this, it was established as main objective to implement an algorithm that has greater accuracy and predictive effectiveness in a module that will be hosted in the INTELLIGENT SYSTEM FOR SUSTAINABLE MOBILITY OF THE UNIVERSITY OF GUAYAQUIL (SIAMS-UG), making use of open Source tools that allow working with Machine Learning to analyze the forecasts of congestion that form around the University of Guayaquil. In order to carry out the project, an exhaustive research was carried out through the methodology of collecting bibliographic citations by means of a systematic review about the defined study area, consequently, for the processing of the information, quantitative data packages were used to obtain the necessary probabilities. As for the development, the cascade methodology was used since being a linear and simple implementation model, it was able to emphasize each phase of the project allowing to have as results after several comparative analyses that the most efficient algorithm was the neural network algorithm, thus proceeding to its implementation in the SIAMS-UG web application in order to meet the stated objective of the problem, which is to obtain improvements in sustainable mobility.
Descripción
PDF
Palabras clave
Análisis, Algoritmos, Movilidad, Redes neuronales, Predicción, Machine Learning., Analysis, Algorithms, Mobility, Neuronal networks, Prediction, Machine Learning