Incremental Market Behavior Classification in Presence of Recurring Concepts
In recent years, the problem of concept drift has gained importance in the financial domain. The succession of manias, panics and crashes have stressed the non-stationary nature and the likelihood of drastic structural or concept changes in the markets. Traditional systems are unable or slow to adap...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2019-01-01
|
| Series: | Entropy |
| Subjects: | |
| Online Access: | http://www.mdpi.com/1099-4300/21/1/25 |
| _version_ | 1798039631172730880 |
|---|---|
| author | Andrés L. Suárez-Cetrulo Alejandro Cervantes David Quintana |
| author_facet | Andrés L. Suárez-Cetrulo Alejandro Cervantes David Quintana |
| author_sort | Andrés L. Suárez-Cetrulo |
| collection | DOAJ |
| description | In recent years, the problem of concept drift has gained importance in the financial domain. The succession of manias, panics and crashes have stressed the non-stationary nature and the likelihood of drastic structural or concept changes in the markets. Traditional systems are unable or slow to adapt to these changes. Ensemble-based systems are widely known for their good results predicting both cyclic and non-stationary data such as stock prices. In this work, we propose RCARF (Recurring Concepts Adaptive Random Forests), an ensemble tree-based online classifier that handles recurring concepts explicitly. The algorithm extends the capabilities of a version of Random Forest for evolving data streams, adding on top a mechanism to store and handle a shared collection of inactive trees, called concept history, which holds memories of the way market operators reacted in similar circumstances. This works in conjunction with a decision strategy that reacts to drift by replacing active trees with the best available alternative: either a previously stored tree from the concept history or a newly trained background tree. Both mechanisms are designed to provide fast reaction times and are thus applicable to high-frequency data. The experimental validation of the algorithm is based on the prediction of price movement directions one second ahead in the SPDR (Standard & Poor’s Depositary Receipts) S&P 500 Exchange-Traded Fund. RCARF is benchmarked against other popular methods from the incremental online machine learning literature and is able to achieve competitive results. |
| first_indexed | 2024-04-11T21:56:32Z |
| format | Article |
| id | doaj.art-2d047ef12fed4d97ac3d5489daf9bbf6 |
| institution | Directory Open Access Journal |
| issn | 1099-4300 |
| language | English |
| last_indexed | 2024-04-11T21:56:32Z |
| publishDate | 2019-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Entropy |
| spelling | doaj.art-2d047ef12fed4d97ac3d5489daf9bbf62022-12-22T04:01:05ZengMDPI AGEntropy1099-43002019-01-012112510.3390/e21010025e21010025Incremental Market Behavior Classification in Presence of Recurring ConceptsAndrés L. Suárez-Cetrulo0Alejandro Cervantes1David Quintana2Department of Computer Science, Universidad Carlos III de Madrid, Leganés, 28911 Madrid, SpainDepartment of Computer Science, Universidad Carlos III de Madrid, Leganés, 28911 Madrid, SpainDepartment of Computer Science, Universidad Carlos III de Madrid, Leganés, 28911 Madrid, SpainIn recent years, the problem of concept drift has gained importance in the financial domain. The succession of manias, panics and crashes have stressed the non-stationary nature and the likelihood of drastic structural or concept changes in the markets. Traditional systems are unable or slow to adapt to these changes. Ensemble-based systems are widely known for their good results predicting both cyclic and non-stationary data such as stock prices. In this work, we propose RCARF (Recurring Concepts Adaptive Random Forests), an ensemble tree-based online classifier that handles recurring concepts explicitly. The algorithm extends the capabilities of a version of Random Forest for evolving data streams, adding on top a mechanism to store and handle a shared collection of inactive trees, called concept history, which holds memories of the way market operators reacted in similar circumstances. This works in conjunction with a decision strategy that reacts to drift by replacing active trees with the best available alternative: either a previously stored tree from the concept history or a newly trained background tree. Both mechanisms are designed to provide fast reaction times and are thus applicable to high-frequency data. The experimental validation of the algorithm is based on the prediction of price movement directions one second ahead in the SPDR (Standard & Poor’s Depositary Receipts) S&P 500 Exchange-Traded Fund. RCARF is benchmarked against other popular methods from the incremental online machine learning literature and is able to achieve competitive results.http://www.mdpi.com/1099-4300/21/1/25ensemble methodsadaptive classifiersrecurrent conceptsconcept driftstock price direction prediction |
| spellingShingle | Andrés L. Suárez-Cetrulo Alejandro Cervantes David Quintana Incremental Market Behavior Classification in Presence of Recurring Concepts Entropy ensemble methods adaptive classifiers recurrent concepts concept drift stock price direction prediction |
| title | Incremental Market Behavior Classification in Presence of Recurring Concepts |
| title_full | Incremental Market Behavior Classification in Presence of Recurring Concepts |
| title_fullStr | Incremental Market Behavior Classification in Presence of Recurring Concepts |
| title_full_unstemmed | Incremental Market Behavior Classification in Presence of Recurring Concepts |
| title_short | Incremental Market Behavior Classification in Presence of Recurring Concepts |
| title_sort | incremental market behavior classification in presence of recurring concepts |
| topic | ensemble methods adaptive classifiers recurrent concepts concept drift stock price direction prediction |
| url | http://www.mdpi.com/1099-4300/21/1/25 |
| work_keys_str_mv | AT andreslsuarezcetrulo incrementalmarketbehaviorclassificationinpresenceofrecurringconcepts AT alejandrocervantes incrementalmarketbehaviorclassificationinpresenceofrecurringconcepts AT davidquintana incrementalmarketbehaviorclassificationinpresenceofrecurringconcepts |