Abstract
Time series forecasts are often influenced by exogenous contextual featuresin addition to their corresponding history. For example, in financial settings,it is hard to accurately predict a stock price without considering publicsentiments and policy decisions in the form of news articles, tweets, etc.Though this is common knowledge, the current state-of-the-art (SOTA)forecasting models fail to incorporate such contextual information, owing toits heterogeneity and multimodal nature. To address this, we introduceContextFormer, a novel plug-and-play method to surgically integrate multimodalcontextual information into existing pre-trained forecasting models.ContextFormer effectively distills forecast-specific information from richmultimodal contexts, including categorical, continuous, time-varying, and eventextual information, to significantly enhance the performance of existing baseforecasters. ContextFormer outperforms SOTA forecasting models by up to 30% ona range of real-world datasets spanning energy, traffic, environmental, andfinancial domains.