Optimal Trend Labeling in Financial Time Series
Predicting asset price trends is often posed as a classification problem, where trends are classified as positive or negative. Since asset price series are noisy and volatile, it is difficult to distinguish true trends from short-term fluctuations. To this end, several trend definitions have been pr...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2023-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10210534/ |
| _version_ | 1797743419696611328 |
|---|---|
| author | Tomislav Kovacevic Andro Mercep Stjepan Begusic Zvonko Kostanjcar |
| author_facet | Tomislav Kovacevic Andro Mercep Stjepan Begusic Zvonko Kostanjcar |
| author_sort | Tomislav Kovacevic |
| collection | DOAJ |
| description | Predicting asset price trends is often posed as a classification problem, where trends are classified as positive or negative. Since asset price series are noisy and volatile, it is difficult to distinguish true trends from short-term fluctuations. To this end, several trend definitions have been proposed in the literature, but it is yet to be known how these trend definitions affect the performance of classification algorithms designed to learn such labels from historical data. In this paper, we define the robustness of the trend labeling algorithm as a measure of how well a classifier designed to learn such labels can withstand a change in the cumulative return considering the classifier’s generalization error. Moreover, we propose a noise model to simulate the desired accuracy score, which allows us to evaluate the robustness of a trend labeling algorithm without the need to train an actual classifier and consequently choose the optimal algorithm in terms of robustness. Experimental results confirm the adequacy of the proposed noise model and show that classification algorithms perform better when trained with such optimal labels. |
| first_indexed | 2024-03-12T14:55:13Z |
| format | Article |
| id | doaj.art-00e8187a9f6947e0ba18a3c9c785205d |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-03-12T14:55:13Z |
| publishDate | 2023-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-00e8187a9f6947e0ba18a3c9c785205d2023-08-14T23:00:47ZengIEEEIEEE Access2169-35362023-01-0111838228383210.1109/ACCESS.2023.330328310210534Optimal Trend Labeling in Financial Time SeriesTomislav Kovacevic0https://orcid.org/0009-0002-1337-1065Andro Mercep1https://orcid.org/0000-0003-1229-3149Stjepan Begusic2https://orcid.org/0000-0002-3186-1749Zvonko Kostanjcar3https://orcid.org/0000-0002-2519-3115Laboratory for Financial and Risk Analytics, Faculty of Electrical Engineering and Computing, University of Zagreb, Zagreb, CroatiaLaboratory for Financial and Risk Analytics, Faculty of Electrical Engineering and Computing, University of Zagreb, Zagreb, CroatiaLaboratory for Financial and Risk Analytics, Faculty of Electrical Engineering and Computing, University of Zagreb, Zagreb, CroatiaLaboratory for Financial and Risk Analytics, Faculty of Electrical Engineering and Computing, University of Zagreb, Zagreb, CroatiaPredicting asset price trends is often posed as a classification problem, where trends are classified as positive or negative. Since asset price series are noisy and volatile, it is difficult to distinguish true trends from short-term fluctuations. To this end, several trend definitions have been proposed in the literature, but it is yet to be known how these trend definitions affect the performance of classification algorithms designed to learn such labels from historical data. In this paper, we define the robustness of the trend labeling algorithm as a measure of how well a classifier designed to learn such labels can withstand a change in the cumulative return considering the classifier’s generalization error. Moreover, we propose a noise model to simulate the desired accuracy score, which allows us to evaluate the robustness of a trend labeling algorithm without the need to train an actual classifier and consequently choose the optimal algorithm in terms of robustness. Experimental results confirm the adequacy of the proposed noise model and show that classification algorithms perform better when trained with such optimal labels.https://ieeexplore.ieee.org/document/10210534/Financial time seriesmachine learningtrend labeling algorithmsoptimization |
| spellingShingle | Tomislav Kovacevic Andro Mercep Stjepan Begusic Zvonko Kostanjcar Optimal Trend Labeling in Financial Time Series IEEE Access Financial time series machine learning trend labeling algorithms optimization |
| title | Optimal Trend Labeling in Financial Time Series |
| title_full | Optimal Trend Labeling in Financial Time Series |
| title_fullStr | Optimal Trend Labeling in Financial Time Series |
| title_full_unstemmed | Optimal Trend Labeling in Financial Time Series |
| title_short | Optimal Trend Labeling in Financial Time Series |
| title_sort | optimal trend labeling in financial time series |
| topic | Financial time series machine learning trend labeling algorithms optimization |
| url | https://ieeexplore.ieee.org/document/10210534/ |
| work_keys_str_mv | AT tomislavkovacevic optimaltrendlabelinginfinancialtimeseries AT andromercep optimaltrendlabelinginfinancialtimeseries AT stjepanbegusic optimaltrendlabelinginfinancialtimeseries AT zvonkokostanjcar optimaltrendlabelinginfinancialtimeseries |