Kavli Affiliate: Wei Gao| Summary:Standard benchmarks evaluate time series foundation models (TSFMs) using aggregate metrics, but these can mask severe failures in critical operating regimes. We introduce regime-stratified evaluation and apply it to three TSFMs on two standard traffic speed benchmarks. Traffic exhibits abrupt regime switching between free-flow and congested states, producing bimodal speed distributions […]
Continue.. Do Time Series Foundation Model Benchmarks Hide Regime-Dependent Failures? Evidence from Traffic Speed Forecasting