Financial Times Series Forecasting of Clustered Stocks
- PDF / 1,610,482 Bytes
- 10 Pages / 595.276 x 790.866 pts Page_size
- 118 Downloads / 220 Views
Financial Times Series Forecasting of Clustered Stocks Felipe Affonso 1
&
Thiago Magela Rodrigues Dias 1
&
Adilson Luiz Pinto 2
# Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract Predicting the stock market is a widely studied field, either due to the curiosity in finding an explanation for the behavior of financial assets or for financial purposes. Among these studies the best techniques use neural networks as a prediction technique. More specifically, the best networks for this purpose are called recurrent neural networks (RNN) and provide an extra option when dealing with a sequence of values. However, a great part of the studies is intended to predict the result of few stocks, therefore, this work aims to predict the behavior of a large number of stocks. For this, similar stocks were grouped based on their correlation and later the algorithm K-means was applied so that similar groups were clustered. After this process, the Long ShortTerm Memory (LSTM) - a type of RNN - was used in order to predict the price of a certain group of assets. Later, predicted prices are compared to the correct prices in order to analyze prices tendency. Results showed that clustering stocks did not influence the effectiveness of the network, once tendency was predicted correct for an average of 48% of time. Investors and portfolio managers can use proposed techniques to simply their daily tasks. Keywords Financial forecasting . Clustering stocks . Artificial neural networks
1 Introduction Forecasting can be defined as the prediction of some future event or events by analyzing the historical data [1]. Times series forecasting plays an imperative role in several realworld problems, such as the financial markets, network traffic, weather forecasting, and petroleum (or oil) industry, among others [2]. As it represents an attempt to estimate a future event based on its past history, the financial market attracts the attention of several researchers interested in this area of knowledge. Some studies argue that it is possible to predict stock market movement [3, 4] once, it is believed that today’s stock price will be dependent on yesterday’s price.
* Felipe Affonso [email protected] Thiago Magela Rodrigues Dias [email protected] Adilson Luiz Pinto [email protected] 1
Centro Federal de Educação Tecnológica de Minas Gerais (CEFET-MG), Belo Horizonte, Minas Gerais, Brazil
2
Universidade Federal de Santa Catarina (UFSC), Florianópolis, Santa Catarina, Brazil
The latest discoveries in machine learning (ML) area shows that algorithms can learn by themselves, recognize patterns, find the best features for a model and many others [5]. By using some deep learning (a ML subarea) method it is possible to better understand financial market than just analyzing stocks prices. Deep Neural Networks have been used several times in order to predict the financial market movement. Studies show that this method is obtaining 48%–54% right predictions [6]. Therefore, it is believed that there is a wide fie
Data Loading...
