Predicting stock price movement using a DBN-RNN
This paper proposes a deep learning-based model to predict stock price movements. The proposed model is composed of a deep belief network (DBN) to learn the latent feature representation from stock prices, and a long short-term memory (LSTM) network to exploit long-range relations within the trading...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Taylor & Francis Group
2021-10-01
|
| Series: | Applied Artificial Intelligence |
| Online Access: | http://dx.doi.org/10.1080/08839514.2021.1942520 |
| _version_ | 1797684819649363968 |
|---|---|
| author | Xiaoci Zhang Naijie Gu Jie Chang Hong Ye |
| author_facet | Xiaoci Zhang Naijie Gu Jie Chang Hong Ye |
| author_sort | Xiaoci Zhang |
| collection | DOAJ |
| description | This paper proposes a deep learning-based model to predict stock price movements. The proposed model is composed of a deep belief network (DBN) to learn the latent feature representation from stock prices, and a long short-term memory (LSTM) network to exploit long-range relations within the trading history. The prediction target of the model is the stock close price direction on the next day. To predict the trend of one stock, the feature of recent trading information is generated from the raw intra-day data through a pre-trained DBN. Then the extracted features are fed into an LSTM classifier to produce the prediction result for the next day. The proposed model was tested on 36 companies in the Shanghai Stock Exchange (SSE) and the Shenzhen Stock Exchange (SZSE), which were selected based on their weights in Chinese A-shares. The experiments cover a span of 12 years, from 2005 to 2016, and the results show that the proposed model offers notable improvements in predicting performance comparing with other learning models. It is also observed that some companies are more predictable than others, which implies that the proposed model can be used for financial portfolio construction. |
| first_indexed | 2024-03-12T00:36:15Z |
| format | Article |
| id | doaj.art-860fad83a7f24fbd823e467fc22e2ed4 |
| institution | Directory Open Access Journal |
| issn | 0883-9514 1087-6545 |
| language | English |
| last_indexed | 2024-03-12T00:36:15Z |
| publishDate | 2021-10-01 |
| publisher | Taylor & Francis Group |
| record_format | Article |
| series | Applied Artificial Intelligence |
| spelling | doaj.art-860fad83a7f24fbd823e467fc22e2ed42023-09-15T09:33:59ZengTaylor & Francis GroupApplied Artificial Intelligence0883-95141087-65452021-10-01351287689210.1080/08839514.2021.19425201942520Predicting stock price movement using a DBN-RNNXiaoci Zhang0Naijie Gu1Jie Chang2Hong Ye3University of Science and Technology of China HefeiUniversity of Science and Technology of China HefeiUniversity of Science and Technology of China HefeiUniversity of Science and Technology of China HefeiThis paper proposes a deep learning-based model to predict stock price movements. The proposed model is composed of a deep belief network (DBN) to learn the latent feature representation from stock prices, and a long short-term memory (LSTM) network to exploit long-range relations within the trading history. The prediction target of the model is the stock close price direction on the next day. To predict the trend of one stock, the feature of recent trading information is generated from the raw intra-day data through a pre-trained DBN. Then the extracted features are fed into an LSTM classifier to produce the prediction result for the next day. The proposed model was tested on 36 companies in the Shanghai Stock Exchange (SSE) and the Shenzhen Stock Exchange (SZSE), which were selected based on their weights in Chinese A-shares. The experiments cover a span of 12 years, from 2005 to 2016, and the results show that the proposed model offers notable improvements in predicting performance comparing with other learning models. It is also observed that some companies are more predictable than others, which implies that the proposed model can be used for financial portfolio construction.http://dx.doi.org/10.1080/08839514.2021.1942520 |
| spellingShingle | Xiaoci Zhang Naijie Gu Jie Chang Hong Ye Predicting stock price movement using a DBN-RNN Applied Artificial Intelligence |
| title | Predicting stock price movement using a DBN-RNN |
| title_full | Predicting stock price movement using a DBN-RNN |
| title_fullStr | Predicting stock price movement using a DBN-RNN |
| title_full_unstemmed | Predicting stock price movement using a DBN-RNN |
| title_short | Predicting stock price movement using a DBN-RNN |
| title_sort | predicting stock price movement using a dbn rnn |
| url | http://dx.doi.org/10.1080/08839514.2021.1942520 |
| work_keys_str_mv | AT xiaocizhang predictingstockpricemovementusingadbnrnn AT naijiegu predictingstockpricemovementusingadbnrnn AT jiechang predictingstockpricemovementusingadbnrnn AT hongye predictingstockpricemovementusingadbnrnn |