Financial time series forecasting using artificial neural networks
Autor(a) principal: | |
---|---|
Data de Publicação: | 2020 |
Tipo de documento: | Dissertação |
Idioma: | eng |
Título da fonte: | Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
Texto Completo: | http://hdl.handle.net/10071/21560 |
Resumo: | This study builds an artificial neural network framework with the use of stacked autoencoders (SAE) to extract deep denoised features, and long short-term memory (LSTM) to generate forecasts for the next-day adjusted closing price of S&P500. Data for seven different stock indices, technical indicators, and macroeconomic variables is used to train three different models: a 'price model' which predicts the next-day price, a 'change model' which predicts the relative change in price, and a ’binary model’ which predicts the probability of a price increase. The models were judged based on predictive accuracy and profitability. Results show the models either fail to generalize well or fall prey to a vicious minimum approximating a naive predictor. Furthermore, the models appear particularly poor at predicting breaks in the series, likely due to their infrequency. This might provide evidence supporting the efficient market hypothesis. |
id |
RCAP_efee78afcb7505cb0a13848012749b73 |
---|---|
oai_identifier_str |
oai:repositorio.iscte-iul.pt:10071/21560 |
network_acronym_str |
RCAP |
network_name_str |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
repository_id_str |
7160 |
spelling |
Financial time series forecasting using artificial neural networksTime series forecastingArtificial neural networksStock marketLong short-term memoryStacked autoencoderThis study builds an artificial neural network framework with the use of stacked autoencoders (SAE) to extract deep denoised features, and long short-term memory (LSTM) to generate forecasts for the next-day adjusted closing price of S&P500. Data for seven different stock indices, technical indicators, and macroeconomic variables is used to train three different models: a 'price model' which predicts the next-day price, a 'change model' which predicts the relative change in price, and a ’binary model’ which predicts the probability of a price increase. The models were judged based on predictive accuracy and profitability. Results show the models either fail to generalize well or fall prey to a vicious minimum approximating a naive predictor. Furthermore, the models appear particularly poor at predicting breaks in the series, likely due to their infrequency. This might provide evidence supporting the efficient market hypothesis.Este estudo constrói modelos de redes neuronais artificiais com o uso de "stacked autoencoders" (SAE) para extrair variáveis latentes sem ruído e "long short-term memory" (LSTM) para gerar previsões para o "next-day adjusted closing price" do S&P500. Dados para sete índices de ações diferentes, indicadores técnicos e variáveis macroeconómicas são usados para treinar três modelos diferentes: um 'modelo de preço' que prevê o preçoo do dia seguinte, um 'modelo de mudança que prevê a mudança relativa no preçoo e um 'modelo binário' que prevê a probabilidade de um aumento de preço. Os modelos foram avaliados com base na sua precisão preditiva e lucratividade. Os resultados mostram que os modelos falham em generalizar bem ou caem num mínimo vicioso que se aproxima de um "naive predictor". Além disso, os modelos parecem particularmente fracos a prever quebras na série, provavelmente devido à sua infrequência. Isto pode fornecer evidências que apoiam a hipótese do mercado eficiente.2021-01-26T14:42:40Z2020-12-09T00:00:00Z2020-12-092020-12info:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/masterThesisapplication/pdfhttp://hdl.handle.net/10071/21560TID:202572749engCardoso, Luís Gil Miguénsinfo:eu-repo/semantics/openAccessreponame:Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)instname:Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informaçãoinstacron:RCAAP2023-11-09T17:25:05Zoai:repositorio.iscte-iul.pt:10071/21560Portal AgregadorONGhttps://www.rcaap.pt/oai/openaireopendoar:71602024-03-19T22:11:22.003766Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) - Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informaçãofalse |
dc.title.none.fl_str_mv |
Financial time series forecasting using artificial neural networks |
title |
Financial time series forecasting using artificial neural networks |
spellingShingle |
Financial time series forecasting using artificial neural networks Cardoso, Luís Gil Miguéns Time series forecasting Artificial neural networks Stock market Long short-term memory Stacked autoencoder |
title_short |
Financial time series forecasting using artificial neural networks |
title_full |
Financial time series forecasting using artificial neural networks |
title_fullStr |
Financial time series forecasting using artificial neural networks |
title_full_unstemmed |
Financial time series forecasting using artificial neural networks |
title_sort |
Financial time series forecasting using artificial neural networks |
author |
Cardoso, Luís Gil Miguéns |
author_facet |
Cardoso, Luís Gil Miguéns |
author_role |
author |
dc.contributor.author.fl_str_mv |
Cardoso, Luís Gil Miguéns |
dc.subject.por.fl_str_mv |
Time series forecasting Artificial neural networks Stock market Long short-term memory Stacked autoencoder |
topic |
Time series forecasting Artificial neural networks Stock market Long short-term memory Stacked autoencoder |
description |
This study builds an artificial neural network framework with the use of stacked autoencoders (SAE) to extract deep denoised features, and long short-term memory (LSTM) to generate forecasts for the next-day adjusted closing price of S&P500. Data for seven different stock indices, technical indicators, and macroeconomic variables is used to train three different models: a 'price model' which predicts the next-day price, a 'change model' which predicts the relative change in price, and a ’binary model’ which predicts the probability of a price increase. The models were judged based on predictive accuracy and profitability. Results show the models either fail to generalize well or fall prey to a vicious minimum approximating a naive predictor. Furthermore, the models appear particularly poor at predicting breaks in the series, likely due to their infrequency. This might provide evidence supporting the efficient market hypothesis. |
publishDate |
2020 |
dc.date.none.fl_str_mv |
2020-12-09T00:00:00Z 2020-12-09 2020-12 2021-01-26T14:42:40Z |
dc.type.status.fl_str_mv |
info:eu-repo/semantics/publishedVersion |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/masterThesis |
format |
masterThesis |
status_str |
publishedVersion |
dc.identifier.uri.fl_str_mv |
http://hdl.handle.net/10071/21560 TID:202572749 |
url |
http://hdl.handle.net/10071/21560 |
identifier_str_mv |
TID:202572749 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.format.none.fl_str_mv |
application/pdf |
dc.source.none.fl_str_mv |
reponame:Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) instname:Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação instacron:RCAAP |
instname_str |
Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação |
instacron_str |
RCAAP |
institution |
RCAAP |
reponame_str |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
collection |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
repository.name.fl_str_mv |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) - Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação |
repository.mail.fl_str_mv |
|
_version_ |
1799134668121964544 |