{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,17]],"date-time":"2026-09-17T14:03:15Z","timestamp":1789653795984,"version":"4.0.0"},"reference-count":36,"publisher":"Wiley","issue":"10","license":[{"start":{"date-parts":[[2026,9,14]],"date-time":"2026-09-14T00:00:00Z","timestamp":1789344000000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/onlinelibrary.wiley.com\/termsAndConditions#vor"},{"start":{"date-parts":[[2026,9,14]],"date-time":"2026-09-14T00:00:00Z","timestamp":1789344000000},"content-version":"tdm","delay-in-days":0,"URL":"http:\/\/doi.wiley.com\/10.1002\/tdm_license_1.1"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["12226352"],"award-info":[{"award-number":["12226352"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["onlinelibrary.wiley.com"],"crossmark-restriction":true},"short-container-title":["Expert Systems"],"published-print":{"date-parts":[[2026,10]]},"abstract":"<jats:title>ABSTRACT<\/jats:title>\n                  <jats:p>\n                    Financial time series prediction faces dual challenges: deep learning models are highly sensitive to hyperparameters, and their vulnerability under extreme market conditions leads to severe performance decay. To address this, this paper proposes the AT\u2010STOA\u2010Transformer\u2010BiLSTM hybrid framework. The framework employs the STOA (Sooty tern optimization algorithm) for two\u2010stage collaborative optimization: first selecting predictive features, then tuning six key hyperparameters of the Transformer\u2010BiLSTM architecture. Concurrently, it integrates an AIC\u2010based adaptive trend decomposition module to separate non\u2010stationary sequences into deterministic trends and stationary residuals. In systematic experiments on China's CSI 300, CSI 500 and CSI 1000 indices, this framework demonstrates significant superiority. Compared to the unoptimized baseline, the entire model reduces the average absolute error (MAE) by approximately 24% on average and increases the coefficient of determination (\n                    <jats:italic>R<\/jats:italic>\n                    <jats:sup>2<\/jats:sup>\n                    ) by 0.043 on average. In most cases, STOA is always superior to or on par with the mainstream meta\u2010heuristic optimizers (COA, DBO, EO, SO and SSA). Crucially, the framework exhibits considerable robustness. During the 2022 Russia\u2013Ukraine conflict\u2014a proxy for extreme market stress\u2014the model maintains stable predictions, significantly outperforming baselines in the decay ratio metric. Furthermore, tests on the Dutch AEX and Swiss SMI indices confirm strong cross\u2010market generalization capabilities. In conclusion, the proposed hybrid framework offers improvements in predictive accuracy, stability and applicability, offering a promising tool for quantitative finance and risk management.\n                  <\/jats:p>","DOI":"10.1111\/exsy.70421","type":"journal-article","created":{"date-parts":[[2026,9,15]],"date-time":"2026-09-15T01:26:52Z","timestamp":1789435612000},"update-policy":"https:\/\/doi.org\/10.1002\/crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Towards Robust Predictions Under Extreme Markets: A Hybrid Framework Integrating Adaptive Trend Decomposition and\n                    <scp>STOA<\/scp>\n                    \u2010Optimized Deep Learning"],"prefix":"10.1111","volume":"43","author":[{"ORCID":"https:\/\/orcid.org\/0009-0002-2599-1712","authenticated-orcid":false,"given":"Shaolun","family":"Jin","sequence":"first","affiliation":[{"name":"School of Mathematics Yangzhou University  Jiangsu China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Qiang","family":"Li","sequence":"additional","affiliation":[{"name":"School of Mathematics Yangzhou University  Jiangsu China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"311","published-online":{"date-parts":[[2026,9,14]]},"reference":[{"key":"e_1_2_10_2_1","first-page":"2546","article-title":"Algorithms for Hyper\u2010Parameter Optimization","volume":"24","author":"Bergstra J.","year":"2011","journal-title":"Advances in Neural Information Processing Systems"},{"key":"e_1_2_10_3_1","doi-asserted-by":"publisher","DOI":"10.1145\/3377930.3389815"},{"key":"e_1_2_10_4_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.frl.2022.102934"},{"key":"e_1_2_10_5_1","volume-title":"Time Series Analysis: Forecasting and Control","author":"Box G. E. P.","year":"2015"},{"key":"e_1_2_10_6_1","doi-asserted-by":"publisher","DOI":"10.1145\/2939672.2939785"},{"key":"e_1_2_10_7_1","doi-asserted-by":"publisher","DOI":"10.3390\/e27101029"},{"key":"e_1_2_10_8_1","first-page":"3","article-title":"STL: A Seasonal\u2010Trend Decomposition Procedure Based on Loess","volume":"6","author":"CLEVELAND RB","year":"1990","journal-title":"Journal of Official Statistics"},{"key":"e_1_2_10_9_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.engappai.2019.03.021"},{"key":"e_1_2_10_10_1","doi-asserted-by":"publisher","DOI":"10.3390\/math9182335"},{"key":"e_1_2_10_11_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.knosys.2019.105190"},{"key":"e_1_2_10_12_1","doi-asserted-by":"publisher","DOI":"10.5281\/zenodo.14985026"},{"key":"e_1_2_10_13_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-05318-5_1"},{"key":"e_1_2_10_14_1","doi-asserted-by":"publisher","DOI":"10.1111\/jtsa.12466"},{"key":"e_1_2_10_15_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.knosys.2022.108320"},{"key":"e_1_2_10_16_1","doi-asserted-by":"publisher","DOI":"10.1162\/neco.1997.9.8.1735"},{"key":"e_1_2_10_17_1","volume-title":"Forecasting: Principles and Practice","author":"Hyndman R. J.","year":"2018"},{"key":"e_1_2_10_18_1","volume-title":"Forecasting: Principles and Practice","author":"Hyndman R. J.","year":"2014"},{"key":"e_1_2_10_19_1","doi-asserted-by":"publisher","DOI":"10.1007\/s10462-023-10567-4"},{"key":"e_1_2_10_20_1","doi-asserted-by":"publisher","DOI":"10.3390\/sci7010007"},{"key":"e_1_2_10_21_1","doi-asserted-by":"publisher","DOI":"10.1287\/mnsc.13.4.B202"},{"key":"e_1_2_10_22_1","doi-asserted-by":"publisher","DOI":"10.1007\/s10618-025-01120-8"},{"key":"e_1_2_10_23_1","doi-asserted-by":"publisher","DOI":"10.1109\/COMITCon.2019.8862255"},{"key":"e_1_2_10_24_1","doi-asserted-by":"publisher","DOI":"10.1109\/5.726791"},{"key":"e_1_2_10_25_1","first-page":"151327","article-title":"Unleashing Foundation Vision Models: Adaptive Transfer for Diverse Data\u2010Limited Scientific Domains","volume":"38","author":"Li Q.","year":"2026","journal-title":"Advances in Neural Information Processing Systems"},{"key":"e_1_2_10_26_1","article-title":"iTransformer: Inverted Transformers Are Effective for Time Series Forecasting arXiv: 2310.06625","author":"Liu Y.","year":"2024","journal-title":"arXiv"},{"key":"e_1_2_10_27_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.advengsoft.2017.07.002"},{"key":"e_1_2_10_28_1","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2026.3683792"},{"key":"e_1_2_10_29_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-95099-5_9"},{"key":"e_1_2_10_30_1","doi-asserted-by":"publisher","DOI":"10.3390\/forecast7040055"},{"key":"e_1_2_10_31_1","doi-asserted-by":"publisher","DOI":"10.1109\/72.97934"},{"key":"e_1_2_10_32_1","first-page":"5998","article-title":"Attention Is All You Need","volume":"30","author":"Vaswani A.","year":"2017","journal-title":"Advances in Neural Information Processing Systems"},{"key":"e_1_2_10_33_1","doi-asserted-by":"publisher","DOI":"10.1007\/s11227-022-04959-6"},{"key":"e_1_2_10_34_1","doi-asserted-by":"publisher","DOI":"10.1109\/TSMCB.2009.2015956"},{"key":"e_1_2_10_35_1","first-page":"73","volume-title":"Proceedings of the 29th Pacific Asia Conference on Language, Information and Computation, Shanghai, China, 30 October \u2010 1 November 2015","author":"Zhang S.","year":"2015"},{"key":"e_1_2_10_36_1","doi-asserted-by":"publisher","DOI":"10.1049\/rpg2.13003"},{"key":"e_1_2_10_37_1","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v35i12.17325"}],"container-title":["Expert Systems"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/pdf\/10.1111\/exsy.70421","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/full-xml\/10.1111\/exsy.70421","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/pdf\/10.1111\/exsy.70421","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,17]],"date-time":"2026-09-17T13:15:29Z","timestamp":1789650929000},"score":1,"resource":{"primary":{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/10.1111\/exsy.70421"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,14]]},"references-count":36,"journal-issue":{"issue":"10","published-print":{"date-parts":[[2026,10]]}},"alternative-id":["10.1111\/exsy.70421"],"URL":"https:\/\/doi.org\/10.1111\/exsy.70421","archive":["Portico"],"relation":{},"ISSN":["0266-4720","1468-0394"],"issn-type":[{"value":"0266-4720","type":"print"},{"value":"1468-0394","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,14]]},"assertion":[{"value":"2026-03-08","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-09-03","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-09-14","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}],"article-number":"e70421"}}