getHillClimbWeights: Hill Climbing Algorithm

In joeornstein/SRP: Stacked Regression and Poststratification

Description

Hill Climbing solution to stacking problem from Caruana et al (2004); use for minimizing LogLoss

Usage

getHillClimbWeights(Y, M, verbose = T)

Arguments

Y	The outcome variable
M	An n x k matrix of out-of-fold predictions of Y, where n is the length of Y, and k is th number of models

Value

The optimal stacking weights

joeornstein/SRP documentation built on Oct. 15, 2020, 8:30 p.m.