Irregular Time Step Forecasting with MAE

Auto-generated stub. Edit this file to add more details.

Applying the Masked Autoencoder reconstruction principle to high-dimensional time series data sampled at irregular intervals without explicit imputation.

Why It Matters

The core technical innovation is adapting the Masked Autoencoder (MAE) strategy, typically for vision/NLP, to directly model and reconstruct time series with missing/irregular data points.

Evidence

leveraging attention mechanisms to reconstruct the entire physical sequence in a single prediction pass.

Related Papers

• openalex-2603.25597-spatiotemporal-system-forecasting-with-irregular-time-steps