Biarchetype analysis for univariate functional data. An application to macroeconomic financial time series
Researchers have introduced a new statistical method called biarchetype analysis, designed for univariate functional data. This technique extends archetype analysis by simultaneously identifying representative structures in both cases (like countries) and temporal aspects. The method expresses cases and time points as mixtures of biarchetypes, offering a more interpretable representation than biclustering by focusing on extreme, representative patterns rather than average centroids. The analysis was applied to 10-year government bond yields from European countries between 2001 and 2025, identifying three distinct time regimes and highlighting Germany, Greece, and Hungary as key country archetypes. AI