Please use this identifier to cite or link to this item: http://bibdigital.epn.edu.ec/handle/15000/22018
Full metadata record
DC FieldValueLanguage
dc.contributor.authorGuerra Espinoza, Esteban Sebastián-
dc.date.accessioned2021-12-27T21:04:18Z-
dc.date.available2021-12-27T21:04:18Z-
dc.date.issued2021-12-27-
dc.identifier.citationGuerra Espinoza, E.S. (2021). Predicción de precios de cierre de acciones en la bolsa de valores mediante técnicas de aprendizaje automático y minería de datos: evaluación de escenarios para transacciones intradía. 99 hojas. Quito : EPN.es_ES
dc.identifier.otherT-FCM/0288/CD 11514-
dc.identifier.urihttp://bibdigital.epn.edu.ec/handle/15000/22018-
dc.descriptionEn el presente trabajo, el problema de predicción de los precios de cierre de las acciones en la bolsa de valores durante el día es estudiado. El problema se considera una tarea muy desafiante debido a la naturaleza volátil de las acciones financieras. Este problema consiste en predecir el precio de cierre de una acción, basándose en información pasada y presente. Con el desarrollo de la Inteligencia Artificial y el aumento de las capacidades computacionales, los métodos de predicción programados han demostrado ser más eficientes para predecir los precios de cierre de las acciones. En este trabajo, se emplean dos técnicas denominadas Redes Neuronales Artificiales y Bosques Aleatorios, para predecir el precio de cierre de las acciones, en ocho bases de datos pertenecientes a diferentes temporalidades del día. Los datos financieros con los que se cuentan son: los precios de apertura (Open), precio máximo (High), precio mínimo (Low) y precio de cierre de las acciones (Close), los cuáles serán utilizados para crear nuevas variables con el fin de encontrar las entradas para el modelo. Los modelos se evalúan utilizando indicadores estratégicos estándar: Error de porcentaje absoluto medio (MAPE) y Error de sesgo medio (MBE). Resultados computacionales de los índices basados en instancias de los modelos para cada una de las bases con las que se cuentan son reportados. Finalmente, conclusiones sobre el presente trabajo son presentadas.es_ES
dc.description.abstractIn the present work, the graph partitioning problem in connected components is studied. The problem consists of partitioning an undirected graph with cost on the edges into a fixed number of connected components, such x x that the number of nodes in each component differs by at most one unit and the total cost of the edges with endnodes in the same subset of nodes that induces a component is minimized. Several integer linear programming models using different approaches (maximizing the edges in the cut or minimizing the edges in the connected components) are presented and the results are compared. Moreover, several families of valid inequalities associated to the polytope of these formulations are exposed, together with a Branch & Cut algorithm for the studied problem. Computational results based on simulated instances of different sizes and densities are reported. Finally, conclusions about the present work are presented. In the present work, the problem of predicting the closing prices of the shares in the stock market during the day is studied. The problem is considered a very challenging task due to the volatile nature of financial stocks. This problem consists of predicting the closing price of a stock, based on past and present information. With the development of Artificial Intelligence and the increase in computational capabilities, programmed prediction methods have proven to be more efficient in predicting closing prices of stocks. In this work, two techniques called Artificial Neural Networks and Random Forests are used to predict the closing price of the shares, in eight databases belonging to different time periods of the day. The financial data that are available are: the opening prices (Open), maximum price (High), minimum price (Low) and closing price of the shares (Close), the results will be used to create new variables with the order to find the inputs for the model. Models are evaluated using standard strategic indicators: Mean Absolute Percentage Error (MAPE) and Mean Bias Error (MBE). Computational results of the indices based on instances of the models for each of the bases that are available are reported. Finally, conclusions about the present work are presented.es_ES
dc.description.sponsorshipGutiérrez Pombosa, Sandra Elizabeth, directores_ES
dc.language.isospaes_ES
dc.publisherQuito, 2021es_ES
dc.rightsopenAccesses_ES
dc.subjectMINERIA DE DATOSes_ES
dc.subjectBOSQUES ALEATORIOSes_ES
dc.titlePredicción de precios de cierre de acciones en la bolsa de valores mediante técnicas de aprendizaje automático y minería de datos: evaluación de escenarios para transacciones intradía.es_ES
dc.typebachelorThesises_ES
Appears in Collections:Tesis Matemáticas (MAT)

Files in This Item:
File Description SizeFormat 
CD 11514.pdf1,17 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.