Marginal Likelihood Estimation with the Cross-Entropy Method

Joshua Chan and Eric Eisenstat (2015)
Econometric Reviews, 34(3), 256-285
[ Journal Version | Working Paper | Code ]

This code computes the marginal likelihood for 6 models: VAR, dynamic factor VAR, TVP-VAR (with constant volatility), probit, logit and t-link models.

The algorithms are based on an adaptive importance sampling method called the improved cross-entropy method developed in Chan and Kroese (2012).

The proposed estimators are demonstrated using two empirical applications involving women's labor market participation and US macroeconomic time series.