Adaptive forecasting improves predictions by customizing model updates
AdaCast: Conditional Parameter Generation for Adaptive Time Series Forecasting
Machine LearningArtificial Intelligence
Summary
Forecasting models that predict future data patterns often use the same approach for all inputs, which can miss important differences between data series. The authors propose AdaCast, a method that creates tailored model adjustments for each input, allowing forecasts to adapt to unique patterns like seasons or trends. This approach works during both training and predicting, resulting in better accuracy across multiple tested datasets. AdaCast also shows improved performance when applied to new, unseen data types, proving its flexibility.
What this means in practice
- •For financial analysts: Produce more accurate forecasts tailored to individual financial time series by adapting model parameters to each dataset's unique patterns.
- •For energy grid operators: Improve demand predictions by dynamically adjusting forecasting models to varying seasonal and temporal behaviors in energy usage data.
Authors
Darahaas Nallagatla, Darryl Cherian Jacob, Pan He
Abstract
Time-series foundation models (TSFMs) have achieved strong forecasting performance across domains. However, most adaptation methods remain static. Existing all-in-one methods learn a single set of dataset-level parameter updates and apply the same adapted model to every input. As a result, they cannot adapt the model parameters to the temporal patterns, seasonality and dynamics of each input time series. This limits their ability to produce forecasts that are tailored to heterogeneous inputs. To address this limitation, we propose AdaCast, a conditional parameter generation framework for time-series forecasting. AdaCast uses a generator to produce input-specific low-rank parameter updates for a frozen pretrained TSFM. These updates adapt the model to each input during both training and inference. Across six public benchmarks, AdaCast consistently outperforms static adaptation baseline in in-domain forecasting and improves zero-shot generalization to held-out datasets across domains. These results demonstrate that conditional parameter generation provides an effective approach for adaptive forecasting.