• JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
 
  Bookmark and Share
 
 
Master's Dissertation
DOI
https://doi.org/10.11606/D.3.2023.tde-03012024-143028
Document
Author
Full name
Thiago Rayam Souza Santos
E-mail
Institute/School/College
Knowledge Area
Date of Defense
Published
São Paulo, 2023
Supervisor
Committee
Costa, Oswaldo Luiz do Valle (President)
Bianchi, Reinaldo Augusto da Costa
Pinto, Afonso de Campos
Title in Portuguese
Sistema de tomada de decisão no mercado de ações utilizando aprendizado de máquina.
Keywords in Portuguese
Aprendizado computacional
Sistema de trading
Abstract in Portuguese
Este trabalho apresenta a aplicação de modelos de aprendizado de máquina baseados em árvores de decisão na avaliação dos momentos ideais para compra e venda de ativos no mercado de ações brasileiro. O aprendizado dos modelos é conduzido utilizando indicadores de mercado, calculados a partir da série histórica de preços. O trabalho apresenta uma aplicação prática, abordando o desafio do tratamento de variáveis não estacionárias, bem como a seleção das melhores variáveis para o modelo. Além de avaliar a capacidade de classificação dos melhores momentos de compra e venda, o estudo também inclui uma análise da aplicação dos modelos na geração de ordens de compra e venda, com a realização do backtesting no período de 2017 até 2023. Os resultados obtidos são comparados com a estratégia conhecida como Moving Average Crossover e uma estratégia baseada em ordens aleatórias, além de serem comparados com o Buy and Hold.
Title in English
Decision-making system in the stock market using machine learning.
Keywords in English
Gradient boosting
Machine learning
Random forest
Trading system
Abstract in English
This work presents the application of machine learning models based on decision trees in evaluating the optimal moments for buying and selling assets in the Brazilian stock market. The learning of these models is conducted using market indicators calculated from the historical price series. The study introduces a practical application, addressing the challenge of handling non-stationary variables, as well as the selection of the best variables for the model. In addition to assessing the classification capability of the best moments to buy and sell, the study also includes an analysis of the models application in generating buy and sell orders, with backtesting conducted from 2017 to 2023. The obtained results are compared with the strategy known as Moving Average Crossover and a strategy based on random orders, in addition to being compared with Buy and Hold.
 
WARNING - Viewing this document is conditioned on your acceptance of the following terms of use:
This document is only for private use for research and teaching activities. Reproduction for commercial use is forbidden. This rights cover the whole data about this document as well as its contents. Any uses or copies of this document in whole or in part must include the author's name.
Publishing Date
2024-01-05
 
WARNING: Learn what derived works are clicking here.
All rights of the thesis/dissertation are from the authors
CeTI-SC/STI
Digital Library of Theses and Dissertations of USP. Copyright © 2001-2024. All rights reserved.