{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,1]],"date-time":"2026-10-01T18:07:21Z","timestamp":1790878041393,"version":"4.1.0"},"reference-count":35,"publisher":"Public Library of Science (PLoS)","issue":"10","license":[{"start":{"date-parts":[[2026,10,1]],"date-time":"2026-10-01T00:00:00Z","timestamp":1790812800000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["www.plosone.org"],"crossmark-restriction":false},"short-container-title":["PLoS One"],"abstract":"<jats:p>The rapid adoption of large language models (LLMs) has sparked growing interest in extending their capabilities beyond traditional natural language tasks, including applications in time series forecasting. This work explores the enhancement of time series forecasting with LLMs by incorporating time-dependent covariates. We propose a set of representative prompting strategies that span a wide range of formats while incorporating covariates, and then implement the proposed framework to evaluate performance across three real-world time series datasets in healthcare, service operations, and transportation. In our experiments, we use GPT-4o-mini and show that incorporating relevant covariate information through an appropriate prompt design can substantially improve forecasting accuracy, reducing prediction errors by up to 70% compared with the no-covariate baseline. Furthermore, the analysis reveals that both the choice of covariate and the prompt structure are critical, as poorly aligned configurations may degrade performance. Finally, we conduct sensitivity analyses to assess the effect of covariate integration in censored settings and quantify uncertainty, confirming that our method achieves statistically significant improvements over existing approaches. These findings underscore the potential of covariate integration in prompt design to bridge the gap between general-purpose LLMs and forecasting tasks.<\/jats:p>","DOI":"10.1371\/journal.pone.0358970","type":"journal-article","created":{"date-parts":[[2026,10,1]],"date-time":"2026-10-01T17:43:46Z","timestamp":1790876626000},"page":"e0358970","update-policy":"https:\/\/doi.org\/10.1371\/journal.pone.corrections_policy","source":"Crossref","is-referenced-by-count":0,"title":["Informed forecasting: Leveraging auxiliary knowledge to boost large language model performance on time series forecasting"],"prefix":"10.1371","volume":"21","author":[{"ORCID":"https:\/\/orcid.org\/0009-0005-2444-1956","authenticated-orcid":true,"given":"Mohammadmahdi","family":"Ghasemloo","sequence":"first","affiliation":[{"name":"Texas A&M University","place":["College Station, Texas, United States of America"],"department":["Department of Industrial and Systems Engineering"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0002-1595-9372","authenticated-orcid":true,"given":"Alireza","family":"Moradi","sequence":"additional","affiliation":[{"name":"Georgia Institute of Technology","place":["Atlanta, Georgia, United States of America"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2026,10,1]]},"reference":[{"key":"pone.0358970.ref001","unstructured":"The Verge. ChatGPT Weekly Users Reach 300 Million in December 2024. 2024. [cited 2025 Jan 22]. https:\/\/www.theverge.com\/2024\/12\/4\/24313097\/chatgpt-300-million-weekly-users"},{"key":"pone.0358970.ref002","unstructured":"Jin M, Wang S, Ma L, Chu Z, Zhang JY, Shi X, et al. Time-llm: Time series forecasting by reprogramming large language models. arXiv preprint arXiv:231001728. 2023."},{"issue":"3","key":"pone.0358970.ref003","doi-asserted-by":"crossref","DOI":"10.1002\/sdr.1773","article-title":"Generative AI and simulation modeling: how should you (not) use large language models like ChatGPT","volume":"40","author":"A Akhavan","year":"2024","journal-title":"Syst Dyn Rev"},{"key":"pone.0358970.ref004","doi-asserted-by":"crossref","first-page":"55682","DOI":"10.1109\/ACCESS.2024.3387941","article-title":"ChatGPT for robotics: Design principles and model abilities","volume":"12","author":"SH Vemprala","year":"2024","journal-title":"IEEE Access"},{"issue":"2","key":"pone.0358970.ref005","doi-asserted-by":"crossref","first-page":"109","DOI":"10.1145\/3715073.3715083","article-title":"Time series forecasting with LLMs: Understanding and enhancing model capabilities","volume":"26","author":"H Tang","year":"2025","journal-title":"SIGKDD Explor Newsl"},{"key":"pone.0358970.ref006","doi-asserted-by":"crossref","unstructured":"Yu X, Chen Z, Ling Y, Dong S, Liu Z, Lu Y. Temporal data meets LLM\u2013explainable financial time series forecasting. arXiv preprint arXiv:230611025. 2023.","DOI":"10.18653\/v1\/2023.emnlp-industry.69"},{"issue":"4","key":"pone.0358970.ref007","doi-asserted-by":"crossref","first-page":"2466","DOI":"10.1007\/s10618-024-01039-6","article-title":"Modeling the impact of out-of-schema questions in task-oriented dialog systems","volume":"38","author":"JA Meem","year":"2024","journal-title":"Data Min Knowl Disc"},{"key":"pone.0358970.ref008","article-title":"Explainable deep learning architecture for multivariate time series forecasting","author":"MA Karimian","year":"2024","journal-title":"Data Min Knowl Discov"},{"issue":"2","key":"pone.0358970.ref009","doi-asserted-by":"crossref","first-page":"2358","DOI":"10.1109\/TPEL.2025.3618228","article-title":"Improved fault diagnosis capability in CHBMCs: Counter design for multiple OC switches via an E-SVM unit","volume":"41","author":"H Lin","year":"2026","journal-title":"IEEE Trans Power Electron"},{"issue":"7","key":"pone.0358970.ref010","doi-asserted-by":"crossref","first-page":"9850","DOI":"10.1109\/TIE.2026.3658639","article-title":"Dissipativity-based multiport stability root-cause identification and mitigation for solid-state transformers","volume":"73","author":"X Meng","year":"2026","journal-title":"IEEE Trans Ind Electron"},{"issue":"11","key":"pone.0358970.ref011","doi-asserted-by":"crossref","first-page":"6851","DOI":"10.1109\/TKDE.2023.3342137","article-title":"PromptCast: A new prompt-based learning paradigm for time series forecasting","volume":"36","author":"H Xue","year":"2024","journal-title":"IEEE Trans Knowl Data Eng"},{"key":"pone.0358970.ref012","unstructured":"Xiao M, Jiang Z, Qian L, Chen Z, He Y, Xu Y, et al. Retrieval-augmented Large Language Models for Financial Time Series Forecasting. arXiv preprint arXiv:250205878. 2025."},{"key":"pone.0358970.ref013","doi-asserted-by":"crossref","first-page":"575","DOI":"10.52783\/jes.1478","article-title":"Forecasting the future: A comprehensive review of time series prediction techniques","volume":"20","author":"MK Sandhya Arora","year":"2024","journal-title":"J Electr Syst"},{"key":"pone.0358970.ref014","doi-asserted-by":"crossref","first-page":"106181","DOI":"10.1016\/j.asoc.2020.106181","article-title":"Financial time series forecasting with deep learning: A systematic literature review: 2005\u20132019","volume":"90","author":"OB Sezer","year":"2020","journal-title":"Appl Soft Comput"},{"issue":"5","key":"pone.0358970.ref015","doi-asserted-by":"crossref","first-page":"118","DOI":"10.18178\/ijmlc.2017.7.5.632","article-title":"Financial time series forecasting \u2013 A deep learning approach","volume":"7","author":"A Dingli","year":"2017","journal-title":"Int J Mach Learn Comput"},{"key":"pone.0358970.ref016","doi-asserted-by":"crossref","first-page":"4","DOI":"10.3389\/fdata.2020.00004","article-title":"AI in healthcare: Time-series forecasting using statistical, neural, and ensemble architectures","volume":"3","author":"S Kaushik","year":"2020","journal-title":"Front Big Data"},{"issue":"3","key":"pone.0358970.ref017","doi-asserted-by":"crossref","DOI":"10.1002\/wene.329","article-title":"Using time series simulation tools for assessing the effects of variable renewable energy generation on power and energy systems","volume":"8","author":"M Koivisto","year":"2018","journal-title":"WIREs Energy Environ"},{"issue":"2","key":"pone.0358970.ref018","doi-asserted-by":"crossref","first-page":"210","DOI":"10.1287\/opre.51.2.210.12780","article-title":"A time-series framework for supply-chain inventory management","volume":"51","author":"Y Aviv","year":"2003","journal-title":"Oper Res"},{"key":"pone.0358970.ref019","doi-asserted-by":"crossref","first-page":"106","DOI":"10.1109\/UKSim.2014.67","article-title":"Stock price prediction using the ARIMA model.","volume-title":"2014 UKSim-AMSS 16th international conference on computer modelling and simulation","author":"AA Ariyo","year":"2014"},{"key":"pone.0358970.ref020","doi-asserted-by":"crossref","DOI":"10.1177\/1847979018808673","article-title":"Forecasting of demand using ARIMA model","volume":"10","author":"J Fattah","year":"2018","journal-title":"Int J Eng Bus Manage"},{"key":"pone.0358970.ref021","doi-asserted-by":"crossref","first-page":"627","DOI":"10.1109\/FSKD.2008.222","article-title":"Forecasting crime using the ARIMA model.","volume-title":"2008 fifth international conference on fuzzy systems and knowledge discovery","author":"P Chen","year":"2008"},{"key":"pone.0358970.ref022","doi-asserted-by":"crossref","first-page":"1104","DOI":"10.1016\/j.rser.2017.02.023","article-title":"A review and analysis of regression and machine learning models on commercial building electricity load forecasting","volume":"73","author":"B Yildiz","year":"2017","journal-title":"Renew Sustain Energy Rev"},{"issue":"9","key":"pone.0358970.ref023","doi-asserted-by":"crossref","first-page":"1413","DOI":"10.1016\/j.energy.2009.06.034","article-title":"Electricity consumption forecasting in Italy using linear regression models","volume":"34","author":"V Bianco","year":"2009","journal-title":"Energy"},{"issue":"2","key":"pone.0358970.ref024","doi-asserted-by":"crossref","first-page":"218","DOI":"10.3390\/rs13020218","article-title":"Random forest regression model for estimation of the growing stock volumes in Georgia, USA, using dense landsat time series and FIA dataset","volume":"13","author":"S Obata","year":"2021","journal-title":"Remote Sens"},{"issue":"5","key":"pone.0358970.ref025","doi-asserted-by":"crossref","first-page":"362","DOI":"10.1080\/07038992.2014.987376","article-title":"Forest monitoring using landsat time series data: A review","volume":"40","author":"A Banskota","year":"2014","journal-title":"Can J Remote Sens"},{"key":"pone.0358970.ref026","doi-asserted-by":"crossref","first-page":"110121","DOI":"10.1016\/j.chaos.2020.110121","article-title":"Deep learning methods for forecasting COVID-19 time-Series data: A Comparative study","volume":"140","author":"A Zeroual","year":"2020","journal-title":"Chaos Solitons Fractals"},{"key":"pone.0358970.ref027","unstructured":"Liang W, Zhang Y, Wu Z, Lepp H, Ji W, Zhao X, et al. Mapping the increasing use of llms in scientific papers. arXiv preprint arXiv:240401268. 2024."},{"key":"pone.0358970.ref028","unstructured":"IBM Research. What are Large Language Models?. 2024. [cited 2025 Jan 22]. Available from: https:\/\/www.ibm.com\/think\/topics\/large-language-models"},{"key":"pone.0358970.ref029","doi-asserted-by":"crossref","DOI":"10.52202\/075280-0861","article-title":"Large language models are zero-shot time series forecasters.","volume-title":"Proceedings of the 37th International Conference on Neural Information Processing Systems. NIPS \u201923","author":"N Gruver","year":"2023"},{"issue":"3","key":"pone.0358970.ref030","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1145\/3719207","article-title":"LLM4TS: Aligning pre-trained LLMs as data-efficient time-series forecasters","volume":"16","author":"C Chang","year":"2025","journal-title":"ACM Trans Intell Syst Technol"},{"key":"pone.0358970.ref031","doi-asserted-by":"crossref","unstructured":"Jia F, Wang K, Zheng Y, Cao D, Liu Y. Gpt4mts: Prompt-based large language model for multimodal time-series forecasting. In: Proceedings of the AAAI Conference on Artificial Intelligence. vol. 38. 2024. pp. 23343\u201351.","DOI":"10.1609\/aaai.v38i21.30383"},{"key":"pone.0358970.ref032","doi-asserted-by":"crossref","unstructured":"Liu C, Xu Q, Miao H, Yang S, Zhang L, Long C, et al. TimeCMA: Towards llm-empowered multivariate time series forecasting via cross-modality alignment. In: Proceedings of the AAAI Conference on Artificial Intelligence. vol. 39. 2025. pp. 18780\u20138.","DOI":"10.1609\/aaai.v39i18.34067"},{"key":"pone.0358970.ref033","unstructured":"Pan Z, Jiang Y, Garg S, Schneider A, Nevmyvaka Y, Song D. S2 IP-LLM: Semantic Space Informed Prompt Learning with LLM for Time Series Forecasting. In: Forty-first International Conference on Machine Learning; 2024."},{"issue":"2","key":"pone.0358970.ref034","doi-asserted-by":"crossref","first-page":"910","DOI":"10.1287\/opre.2021.0597","article-title":"A doubly stochastic simulator with applications in arrivals modeling and simulation","volume":"73","author":"Y Zheng","year":"2025","journal-title":"Oper Res"},{"issue":"4","key":"pone.0358970.ref035","doi-asserted-by":"crossref","first-page":"679","DOI":"10.1016\/j.ijforecast.2006.03.001","article-title":"Another look at measures of forecast accuracy","volume":"22","author":"RJ Hyndman","year":"2006","journal-title":"Int J Forecast"}],"container-title":["PLOS One"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0358970","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,1]],"date-time":"2026-10-01T17:43:53Z","timestamp":1790876633000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0358970"}},"subtitle":[],"editor":[{"given":"Dinesh Kumar","family":"Nishad","sequence":"first","affiliation":[{"name":"Dr Shakuntala Misra National Rehabilitation University","place":["INDIA"]}],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2026,10,1]]},"references-count":35,"journal-issue":{"issue":"10","published-online":{"date-parts":[[2026,10,1]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pone.0358970","relation":{},"ISSN":["1932-6203"],"issn-type":[{"value":"1932-6203","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,10,1]]}}}