convergence_plot
Can be used to check model convergence.
convergence_plot(model, Xtest, ytest, burnin = 0)
model |
list containing a model of class "hs_rulefit". |
Xtest |
Out of bag sample to check error. |
ytest |
response of test data. |
burnin |
Number of samples disregarded as burnin. |
Convergence is checked by the convergence of the prediction error on unseen test data, to find a suitable number of iterations, in the spirit of gradient boosting. To check convergence on the Training data just use training X and y instead of Xtest and ytest.
library(MASS)
data(Boston)
#Split in train and test data
N = nrow(Boston)
train = sample(1:N, 400)
Xtrain = Boston[train,-14]
ytrain = Boston[train, 14]
Xtest = Boston[-train, -14]
ytest = Boston[-train, 14]
hrres = HorseRuleFit(X = Xtrain, y=ytrain,
thin=1, niter=100, burnin=10,
L=5, S=6, ensemble = "both", mix=0.3, ntree=100,
intercept=FALSE, linterms=1:13, ytransform = "log",
alpha=1, beta=2, linp = 1, restricted = 0)
#Check the model convergence out of sample
convergence_plot(hrres, Xtest, ytest, burnin = 10)Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.