AI RESEARCH

Global and Local Topology-Aware Attention with Persistent Homology and Euler Biases for Time-Series Forecasting

arXiv CS.LG

ArXi:2605.03163v1 Announce Type: new Scientific time series often encode predictive geometric structure, including connectivity, cycles, shell-like geometry, directional changes, and nonlinear neighborhoods, that standard dot-product attention does not explicitly represent. We We evaluate guarded topology-aware variants across three architecture families: lightweight attention/Ridge, PatchTSTForRegression, and TimeSeriesTransformerForPrediction.