{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,10,11]],"date-time":"2025-10-11T04:42:00Z","timestamp":1760157720989,"version":"build-2065373602"},"reference-count":20,"publisher":"MDPI AG","issue":"10","license":[{"start":{"date-parts":[[2025,10,9]],"date-time":"2025-10-09T00:00:00Z","timestamp":1759968000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"the Steel Manufacturing Simulation and Visualization Consortium","award":["DE-EE0009390","DE-AC05-00OR22725"],"award-info":[{"award-number":["DE-EE0009390","DE-AC05-00OR22725"]}]},{"DOI":"10.13039\/100000015","name":"the US Department of Energy\u2019s Office of Energy Efficiency and Renewable Energy under the Industrial Technologies Office","doi-asserted-by":"publisher","award":["DE-EE0009390","DE-AC05-00OR22725"],"award-info":[{"award-number":["DE-EE0009390","DE-AC05-00OR22725"]}],"id":[{"id":"10.13039\/100000015","id-type":"DOI","asserted-by":"publisher"}]},{"name":"UT-Battelle, LLC","award":["DE-EE0009390","DE-AC05-00OR22725"],"award-info":[{"award-number":["DE-EE0009390","DE-AC05-00OR22725"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Data"],"abstract":"<jats:p>The process of time-series forecasting such as predicting trajectories of silicon content in blast furnaces is a difficult task. Most time-series approaches today focus on scalar-type MSE loss optimization. This optimization approach, while widely common, could benefit from the use of human expert or process-level preferences. In this paper, we introduce a novel alignment and fine-tuning approach that involves learning from a corpus of preferred and dis-preferred time-series prediction trajectories. Our contributions include (1) a preference annotation pipeline for time-series forecasts, (2) the application of Score-based Preference Optimization (SPO) to train decoder-only transformers from preferences, and (3) results showing improvements in forecast quality. The approach is validated on both proprietary blast furnace data and the UCI Appliances Energy dataset. The proposed preference corpus and training strategy offer a new option for fine-tuning sequence models in industrial settings.<\/jats:p>","DOI":"10.3390\/data10100161","type":"journal-article","created":{"date-parts":[[2025,10,9]],"date-time":"2025-10-09T09:41:02Z","timestamp":1760002862000},"page":"161","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["A Preferences Corpus and Annotation Scheme for Human-Guided Alignment of Time-Series GPTs"],"prefix":"10.3390","volume":"10","author":[{"given":"Ricardo A.","family":"Calix","sequence":"first","affiliation":[{"name":"Department of Computer Information Technology, Purdue University Northwest, Hammond, IN 46323, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Tyamo","family":"Okosun","sequence":"additional","affiliation":[{"name":"Center for Innovation Through Visualization and Simulation (CIVS) and Steel Manufacturing Simulation and Visualization Consortium (SMSVC), Purdue University Northwest, Hammond, IN 46323, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Chenn","family":"Zhou","sequence":"additional","affiliation":[{"name":"Center for Innovation Through Visualization and Simulation (CIVS) and Steel Manufacturing Simulation and Visualization Consortium (SMSVC), Purdue University Northwest, Hammond, IN 46323, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Hong","family":"Wang","sequence":"additional","affiliation":[{"name":"Oak Ridge National Laboratory, Oak Ridge, TN 37830, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2025,10,9]]},"reference":[{"key":"ref_1","unstructured":"Radford, A., Narasimhan, K., Salimans, T., and Sutskever, I. (2018). Improving Language Understanding by Generative Pre-Training. OpenAI Blog, 1, Available online: https:\/\/www.mikecaptain.com\/resources\/pdf\/GPT-1.pdf."},{"key":"ref_2","doi-asserted-by":"crossref","unstructured":"Zhou, H., Zhang, S., Peng, J., Zhang, S., Li, J., Xiong, H., and Zhang, W. (2021, January 2\u20139). Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting. Proceedings of the AAAI Conference on Artificial Intelligence, Virtual.","DOI":"10.1609\/aaai.v35i12.17325"},{"key":"ref_3","unstructured":"Wu, H., Xu, J., Wang, J., and Long, M. (2021, January 6\u201314). Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting. Proceedings of the Advances in Neural Information Processing Systems, Online."},{"key":"ref_4","unstructured":"Nie, Y., Nguyen, N., Sinthong, P., and Kalagnanam, J. (2023, January 1\u20135). A Time Series is Worth 64 Words: Long-term Forecasting with Transformers. Proceedings of the International Conference on Learning Representations (ICLR), Kigali, Rwanda."},{"key":"ref_5","first-page":"60","article-title":"LLM4TS: Aligning Pre-Trained LLMs as Data-Efficient Time-Series Forecasters","volume":"16","author":"Ching","year":"2025","journal-title":"Acm Trans. Intell. Syst. Technol."},{"key":"ref_6","unstructured":"Niu, W., Xie, Z., Sun, Y., He, W., Xu, M., and Hao, C. (2025). LangTime: A Language-Guided Unified Model for Time Series Forecasting with Proximal Policy Optimization. arXiv."},{"key":"ref_7","first-page":"3","article-title":"Chaotic identification and prediction of silicon content in hot metal","volume":"12","author":"Gao","year":"2005","journal-title":"J. Iron Steel Res. Int."},{"key":"ref_8","doi-asserted-by":"crossref","unstructured":"Zhou, K., Fei, H., and Gong, J. (2021, January 22\u201324). Prediction of Silicon Content of Molten Iron in Blast Furnace Based on Particle Swarm - Random Forest. Proceedings of the 2021 33rd Chinese Control and Decision Conference (CCDC), Kunming, China.","DOI":"10.1109\/CCDC52312.2021.9602093"},{"key":"ref_9","first-page":"30","article-title":"Application of fuzzy Bayesian network to prediction of silicon content in molten iron of blast furnace","volume":"29","author":"Liu","year":"2005","journal-title":"Metall. Autom."},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"1490","DOI":"10.7498\/aps.54.1490","article-title":"Chaotic analysis for blast furnace ironmaking process","volume":"54","author":"Gao","year":"2005","journal-title":"Acta Phys. Sin."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"316","DOI":"10.2355\/isijinternational.42.316","article-title":"Application of nonlinear time series analysis to the prediction of silicon content of pig iron","volume":"42","author":"Waller","year":"2002","journal-title":"ISIJ Int."},{"key":"ref_12","unstructured":"Christiano, P.F., Leike, J., Brown, T.B., Martic, M., Legg, S., and Amodei, D. (2017, January 4\u20139). Deep reinforcement learning from human preferences. Proceedings of the Advances in Neural Information Processing Systems, Long Beach, CA, USA."},{"key":"ref_13","unstructured":"Ouyang, L., Wu, J., Jiang, X., Almeida, D., Wainwright, C., Mishkin, P., Zhang, C., Agarwal, S., Slama, K., and Ray, A. (2022). Training language models to follow instructions with human feedback. arXiv."},{"key":"ref_14","unstructured":"Rafailov, R., Sharma, A., Mitchell, E., Ermon, S., and Manning, C.D. (2023). Direct Preference Optimization: Your Language Model is Secretly a Reward Model. arXiv."},{"key":"ref_15","unstructured":"Rafailov, R., Sharma, A., Mitchell, E., Manning, C.D., Ermon, S., and Finn, C. (2023, January 10\u201316). Direct Preference Optimization: Your Language Model is Secretly a Reward Model. Proceedings of the 36th Conference on Neural Information Processing Systems (NeurIPS), New Orleans, LA, USA."},{"key":"ref_16","unstructured":"(2024). DeepSeek. DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models. arXiv."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"81","DOI":"10.1016\/j.enbuild.2017.01.083","article-title":"Data Driven Prediction Models of Energy Use of Appliances in a Low-Energy House","volume":"140","author":"Candanedo","year":"2017","journal-title":"Energy Build."},{"key":"ref_18","unstructured":"Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A.N., Kaiser, l., and Polosukhin, I. (2017, January 4\u20139). Attention is All You Need. Proceedings of the 31st International Conference on Neural Information Processing Systems (NeurIPS), Long Beach, CA, USA."},{"key":"ref_19","first-page":"324","article-title":"Rank Analysis of Incomplete Block Designs: I. The Method of Paired Comparisons","volume":"39","author":"Bradley","year":"1952","journal-title":"Biometrika"},{"key":"ref_20","unstructured":"Calix, R. (2025, September 28). Preferences and Time Series GPT GitHub Data and Code. Available online: https:\/\/github.com\/rcalix1\/PreferencesTimeSeriesGPT."}],"container-title":["Data"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2306-5729\/10\/10\/161\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,11]],"date-time":"2025-10-11T04:15:48Z","timestamp":1760156148000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2306-5729\/10\/10\/161"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,10,9]]},"references-count":20,"journal-issue":{"issue":"10","published-online":{"date-parts":[[2025,10]]}},"alternative-id":["data10100161"],"URL":"https:\/\/doi.org\/10.3390\/data10100161","relation":{},"ISSN":["2306-5729"],"issn-type":[{"type":"electronic","value":"2306-5729"}],"subject":[],"published":{"date-parts":[[2025,10,9]]}}}