Deep generative models provide an effective modern solution to alleviating the problem of data scarcity by synthesizing realistic time series. However, despite the considerable progress achieved by existing deep time series generators, two key challenges remain: (1) learning hierarchical representations for heterogeneo...
Chun-Yi Hou, Xiangfei Qiu, Han-Yin Cheng et al.· 0 citations
XMatch (EXogenous MATCHing), a covariate-aware forecasting model that realizes a strategy that matches future and historical exogenous patterns and uses the corresponding endogenous patterns to enhance forecasting through a tree-structured matching process that adaptively adjusts the number of exogenous variables used...
Zi-Yang Zhang, Han-Yin Cheng, Xiangfei Qiu et al.· 0 citations
The proposed QiYao-M is a role-aware multimodal TSFM that models the two types of modalities separately, including endogenous and exogenous modalities, and introduces Endo-Modality Proxy Training to train this retrieval module without exogenous multimodal pretraining data.
Han-Yin Cheng, Lin-Feng Wang, Zheng-Bo Qu et al.· 2 citations
HALO is proposed, which enhances time series generation via Hyperspherical Latents and Masked Autoregressive modeling to achieve this goal by tackling two key bottlenecks: variance and scale heterogeneity of continuous latent representations and the difficulty of balancing generation efficiency with temporal correlatio...
In this work, we introduce FLAME, a family of extremely lightweight and capable Time Series Foundation Models, which support versatile forecasting tasks via generative probabilistic modeling, while ensuring both efficiency and robustness. FLAME utilizes the Legendre Memory for strong generalization capabilities. By ada...
This work proposes WorldTS, a world-modeling based forecasting framework that integrates multimodal covariates directly into the forecasting to further improve forecasting performance and employs a two-stage training strategy.
A novel pattern-guided mixture-of-experts that expands model capacity through sparse activation and uses shallow patch similarities to constrain deep-layer routing, guiding expert specialization across heterogeneous time series and an implicit quantile network head that predicts arbitrary quantiles to characterize pred...