arXiv Machine Learning

CoRe: Coherence and Relational Alignment for Multivariate Time Series Forecasting

CoRe is a model‑agnostic learning objective for direct multivariate time‑series forecasting that replaces pointwise errors with two output‑space constraints: a frequency coherence loss aligning predicted and target spectra, and a low‑rank relational graph loss matching pairwise differences in a PCA subspace. The objective introduces no trainable parameters and can be applied to existing forecasting backbones by changing only the loss. Experiments on standard benchmarks show that CoRe improves strong baselines, compares favorably with recent forecasting objectives, and remains effective across different backbones, datasets, and hyperparameter settings.

arXiv Machine Learning
Oct 9

SteerCast: Retrieval-Based Latent Steering for Decoder-Only Time Series Forecasting

SteerCast is a retrieval‑based latent steering technique that enhances decoder‑only time‑series forecasters during inference without updating model parameters. It builds a database from the training set, storing each history window’s representation and a steering vector that captures the difference between the model’s prediction and the ground‑truth continuation in latent space. During testing, SteerCast retrieves nearest neighbors, aggregates their steering vectors, and injects the resulting signal into the forecaster’s hidden states at each autoregressive step, guiding predictions toward trajectories seen in similar training cases. Experiments on diverse multivariate benchmarks and multiple horizons demonstrate consistent accuracy improvements over fine‑tuned backbones and other retrieval‑based baselines, all while requiring no additional training beyond the original fine‑tuning and using only the training set as a retrieval corpus.

By Van Dai Do, Huu Hiep Nguyen, Minh Hoang Nguyen, Hung Le
arXiv AI
Oct 9

Timer-M1: A Multivariate Time Series Foundation Model via Learning Primitives

Timer‑M1 is a pretrained multivariate time‑series foundation model that learns with primitives for zero‑shot forecasting. It uses a primitive‑based data synthesis and pretraining pipeline to generate series with shared temporal and relational patterns, assembling them into multivariate samples with distinct channel roles. The model employs gated two‑dimensional Transformer blocks that dynamically allocate cross‑variate attention, achieving top performance on several large‑scale forecasting benchmarks.

By Haoran Zhang, Haixuan Liu, Xingjian Su, Yong Liu, Zhi Chen, Yuxuan Wang, Jianmin Wang, Mingsheng Long
arXiv Machine Learning
Oct 28

Model-agnostic predictability quantification of multivariable time series

The paper introduces the Multivariable Predictability Score (MAPS), a model‑agnostic metric designed to quantify the intrinsic predictability of multivariable time‑series forecasting tasks. MAPS leverages covariate information across various lookback windows and forecasting horizons to assess predictive information, addressing limitations of existing measures that often miss nonlinear dependencies or focus on univariate series. Experiments on synthetic and real‑world datasets show MAPS aligns closely with expected reference predictability (r = 0.963) and provides more reliable projections than six baseline measures.

By Yuxuan Shu, Vasileios Lampos
arXiv Machine Learning
Oct 16

Learning to Forecast on Multiple Spatiotemporal Hierarchies

The paper introduces a graph deep learning framework that learns to forecast spatiotemporal data across multiple hierarchical structures, using variations among hierarchies as multi‑task regularization. By jointly training on different aggregation levels, the model reduces over‑specialization and improves forecasting accuracy. A novel top‑down operator ensures coherent forecasts with linear time and space complexity, making the approach computationally efficient.

By Valentina Moretti, Cesare Alippi, Andrea Cini
arXiv Machine Learning
Sep 30

HALO: Enhancing Time Series Generation via Hyperspherical Latents and Masked AutoregRessive Modeling

HALO introduces a hyperspherical VAE to constrain continuous latent representations to a fixed‑radius shell, stabilizing numerical fluctuations. It then employs a masked autoregressive model that balances parallel decoding with temporal correlation learning, reducing inference steps and improving stability. Experiments show HALO achieves state‑of‑the‑art generation performance with significantly better inference efficiency compared to existing baselines.

By Chunyi Hou, Xiangfei Qiu, Hanyin Cheng, Yutong Li, Bin Yang