← Back to all news
arXiv AI By Seunghan Lee, Sangjun Han, Jun Seo, Junhyeok Kang, Jaehoon Lee, Tae Yoon Lim, Dongwan Kang, Hwanil Choi, Minjae Kim, Sungdong Yoo, Soonyoung Lee, Wonbin Ahn

EXAONE Demand 1.0: A Time Series Foundation Model for Demand Forecasting

Read the original on arXiv AI →

EXAONE Demand 1.0 is a time‑series foundation model specifically designed for demand forecasting. It is built on a large, demand‑focused corpus of 11.3 million series and 48.4 billion observations, supplemented by a synthetic generator to capture under‑represented demand behaviors. The model uses a demand‑aware adapter with low‑rank branches for four demand classes and a router that weighs each branch based on eight scale‑free statistics, achieving superior performance over 36 existing TSFMs on 22 held‑out datasets.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

Related stories

arXiv Machine Learning
Jul 21

Time-Aware Prior Fitted Networks for Zero-Shot Forecasting with Exogenous Variables

arXiv:2603. 15802v2 Announce Type: replace Abstract: In many time series forecasting settings, the target time series is accompanied by exogenous covariates, such as promotions and prices in retail demand; temperature in energy load; calendar and holiday indicators for traffic or sales; and grid load or fuel costs in electricity pricing.

By Andres Potapczynski, Ravi Kiran Selvam, Tatiana Konstantinova, Malcolm Wolff, Kin G. Olivares, Ruijun Ma, Michael W. Mahoney, Andrew Gordon Wilson, Boris N. Oreshkin, Dmitry Efimov
benchmarks
More like this →
arXiv AI
Sep 18

QUALS: Corpus Equilibrium for Universal Forecasting via Pattern Quantization and Learnability Synchronization

QUALS is a large‑scale time‑series corpus equilibrium framework designed to improve data efficiency for zero‑shot forecasting. It uses pattern quantization to decode heterogeneous patterns from mixed corpora and a learnability synchronization mechanism to calibrate sampling weights, bridging the optimization gap between simple and complex motifs. Benchmarks show that pre‑training on QUALS yields superior zero‑shot performance even with reduced training budgets.

By Yujie Li, Zezhi Shao, Chengqing Yu, Yisong Fu, Weijie Zhu, Yifan Du, Jilin Hu, Bin Yang, Yongjun Xu, Fei Wang
efficiencybenchmarks
More like this →
Hugging Face Trending Papers
Jun 25

How Good Can Linear Models Be for Time-Series Forecasting?

Time-series forecasting research has been moving steadily toward larger architectures, from specialized transformers to general-purpose foundation models, on the assumption that capacity is what unlocks accuracy. We take the opposite position: most of the gap can be closed at far lower cost by tuning preprocessing rather than scaling models.

llmsbenchmarks
More like this →