comparePerformance: Compare MCP, SCAD, LASSO, and BCVO for high dimensional...
In JieGroup/bc: Model Selection with Bridge Criterion

Description Usage Arguments Value

This function compares the performance of MCP, SCAD, LASSO, and BCVO on on a single dataset, in terms of prediction loss and over/under-fitting.

1
2
3

comparePerformance(dat, dat_test, hd_methods, dfmax, ratio, weight_method,
  criteria = c("GIC2", "AIC", "BIC", "BC"), penaltyBC = NULL,
  adaptiveBC = TRUE, methodPI = "BC")

`dat`	Matrix of design that is used for training
`dat_test`	Matrix of design that is used for comparing performance
`hd_methods`	Vector of strings ("MCP", "SCAD", "lasso") indicating which penalized regression methods are used to generate initial candidate subsets
`dfmax`	Integer of maximum number of variables selected by each penalized regression
`ratio`	Value (0 to 1) of proportion of data in obtaining initial candidate subsets
`weight_method`	String indicating which method to use for weighting
`criteria`	Vector of strings ('GIC2, GICn, Cp, AIC, BIC, BC') indicating criteria to use in selecting the final subset
`penaltyBC`	Vector of non-default penalty values in using BC, default to NULL so that only the suggested value n^(1/3) will be considered
`adaptiveBC`	Boolean indicating whether to use adaptive BC, default to FALSE so that only the suggested value n^(1/3) will be considered
`methodPI`	String ('BC', 'Drop', or 'Both') indicating the method to calculate BC, default to 'BC'

List of prediction loss, prediction correlation, number of overfitted variables, number of underfitted variables, prediction residual, and selected subsets

JieGroup/bc documentation built on June 1, 2019, 12:48 p.m.