Binary Generalized Extreme Value Additive Modelling


bgeva provides a function for univariate modelling for binary rare events data with linear and nonlinear predictor effects when using the quantile function of the Generalized Extreme Value random variable.


bgeva provides a function for flexible regression models for binary rare events data. The underlying representation and estimation of the model is based on a penalized regression spline approach, with automatic smoothness selection. The numerical routine carries out function minimization using a trust region algorithm from the package trust in combination with an adaptation of a low level smoothness selection fitting procedure from the package mgcv.

bgeva supports the use of many smoothers as extracted from mgcv. Scale invariant tensor product smooths are not currently supported. Estimation is by penalized maximum likelihood with automatic smoothness selection achieved by using the approximate Un-Biased Risk Estimator (UBRE).

Confidence intervals for smooth components are derived using a Bayesian approach. Approximate p-values for testing individual smooth terms for equality to the zero function are also provided. Functions plot.bgeva and summary.bgeva extract such information from a fitted bgevaObject. Variable selection is also possible via the use of shrinakge smoothers or information criteria.


Raffaella Calabrese (University of Milano-Bicocca, Department of Statistics and Quantitative Methods), Giampiero Marra (University College London, Department of Statistical Science) and Silvia Osmetti (University Cattolica del Sacro Cuore, Department of Statistics)

Maintainer: Giampiero Marra


Calabrese R., Marra G., Osmetti S.A. (2013), Bankruptcy Prediction of Small and Medium Enterprises Using a Flexible Binary Generalized Extreme Value Model. Submitted.

Calabrese R., Osmetti S.A. (2013), Modelling SME Loan Defaults as Rare Events: The Generalized Extreme Value Regression Model. Journal of Applied Statistics.

See Also


comments powered by Disqus