Título: An approach using Artificial Neural Network and Genetic Algorithm for Day Trade Portfolio Selection
Autores: Paula Campigotto and Omir Correia Alves Junior.
Resumo:
In the financial market there are several types of investors, from the most conservative to the most daring, who are subject to greater risks in the expectation of greater returns on their investments. However, the concept of risk, in investment portfolios, makes it possible to measure it in different ways. This paper aims to present a method created to select portfolios for Day Trade financial investments using different metric risks, such as CVaR, EWMA and GARCH, and the ensemble of Genetic Algorithm NSGA-II and LSTM Artificial Neural Network, comparing it’s selected portfolios’ performance with another method which uses only NSGA-II and Buy and Hold financial strategy. The results show that the proposed method, with LSTM ANN achieved better returns in the year of 2019.
Palavras-chave:
Portfolio Selection, Genetic Algorithm, Artificial Neural Network.
Páginas: 7
Código DOI: 10.21528/CBIC2021-88
Artigo em pdf: CBIC_2021_paper_88.pdf
Arquivo BibTeX: CBIC_2021_88.bib