Problem
Research questions and friction points this paper is trying to address.
multivariate time series forecasting
tabular foundation models
inter-channel interactions
zero-shot forecasting
Zero-shot Multivariate Time Series Forecasting Using Tabular Prior Fitted Networks
🤖 AI Summary
This work proposes the first general-purpose zero-shot forecasting framework tailored for multivariate time series, addressing the limitation of existing tabular foundation model–based approaches that typically decompose the problem into independent univariate tasks and thereby neglect dynamic inter-variable dependencies. By reframing forecasting as a sequence of scalar regression problems, the method leverages tabular foundation models with inherent regression capabilities—such as TabPFN-TS—to explicitly model interactions among variables. This approach abandons the restrictive independence assumption prevalent in prior methods and demonstrates substantial performance gains over current state-of-the-art tabular techniques across multiple benchmark datasets, establishing its effectiveness and competitiveness in zero-shot settings.
📝 Abstract
Tabular foundation models, particularly Prior-data Fitted Networks like TabPFN have emerged as the leading contender in a myriad of tasks ranging from data imputation to label prediction on the tabular data format surpassing the historical successes of tree-based models. This has led to investigations on their applicability to forecasting time series data which can be formulated as a tabular problem. While recent work to this end has displayed positive results, most works have limited their treatment of multivariate time series problems to several independent univariate time series forecasting subproblems, thus ignoring any inter-channel interactions. Overcoming this limitation, we introduce a generally applicable framework for multivariate time series forecasting using tabular foundation models. We achieve this by recasting the multivariate time series forecasting problem as a series of scalar regression problems which can then be solved zero-shot by any tabular foundation model with regression capabilities. We present results of our method using the TabPFN-TS backbone and compare performance with the current state of the art tabular methods.
Innovation
Methods, ideas, or system contributions that make the work stand out.
zero-shot forecasting
multivariate time series
tabular foundation models
TabPFN
inter-channel interactions
🔎 Similar Papers
No similar papers found.
