ENTROPY-BASED SEEMINGLY UNRELATED REGRESSION
No hay miniatura disponible
Fecha
Título de la revista
ISSN de la revista
Título del volumen
Editor
Resumen
Descripción
We adapt the classical SUR procedure to a minimum cross entropy approach to estimate linear systems of equations where the errors across equations are correlated. We conclude that our entropy-based approach may provide a reasonable substitute for SUR in cases where classical methods may not be applied due to shortages of data.