| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 7.82 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
Este trabalho apresenta uma análise sobre a previsão de preços de ações no mercado financeiro, com ênfase em abordagens baseadas em modelos de séries temporais e técnicas de Deep Learning. Foram explorados conceitos fundamentais de análise técnica, como médias exponenciais e simples, e analisados diversos índices globais, para serem utilizados como entradas para os modelos de Machine Learning, que incluem
Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), Convolutional Neural Network (CNN) e XGBoost. Tudo isto objetivo de capturar dependências temporais nos dados, permitindo assim previsões mais precisas dos movimentos futuros do mercado. Ao abordar esta questão, Foi enfatizada a importância do pré-processamento adequado dos dados, incluindo limpeza, normalização e seleção das features relevantes. Além disso, foi discutida a necessidade da avaliação do desempenho dos modelos através de métricas adequadas e técnicas de validação para garantir resultados confiáveis.
This paper presents an analysis of stock price forecasting in the financial market, with an emphasis on approaches based on time series models and Deep Learning techniques. Fundamental concepts of technical analysis were explored, such as exponential and simple averages, and various global indices were analyzed to be used as inputs for Machine Learning models, including Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), Convolutional Neural Network (CNN) and XGBoost. All of this is aimed at capturing temporal dependencies in the data, thus enabling more accurate forecasts of future market movements. In addressing this issue, the importance of proper data pre-processing was emphasized, including cleaning, normalizing and selecting the relevant features. In addition, the need to evaluate model performance using appropriate metrics and validation techniques to ensure reliable results was discussed.
This paper presents an analysis of stock price forecasting in the financial market, with an emphasis on approaches based on time series models and Deep Learning techniques. Fundamental concepts of technical analysis were explored, such as exponential and simple averages, and various global indices were analyzed to be used as inputs for Machine Learning models, including Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), Convolutional Neural Network (CNN) and XGBoost. All of this is aimed at capturing temporal dependencies in the data, thus enabling more accurate forecasts of future market movements. In addressing this issue, the importance of proper data pre-processing was emphasized, including cleaning, normalizing and selecting the relevant features. In addition, the need to evaluate model performance using appropriate metrics and validation techniques to ensure reliable results was discussed.
Description
Keywords
Machine learning Predictions Stock market Neural networks Previsões Mercado de ações Redes neuronais