Booster
library(unifiedml) # this package
base_learner <- function() {
Model$new(lm)
}
boost_model <- BoostingModel$new(
base_learner_fn = base_learner,
B = 100,
eta = 0.1
)
# Generate sample data
set.seed(123)
X <- matrix(rnorm(200), ncol = 4)
y <- 2*X[,1] - 1.5*X[,2] + rnorm(50)
# Fit and predict
boost_model$fit(X, y)
## Iteration 10 / 100
## Iteration 20 / 100
## Iteration 30 / 100
## Iteration 40 / 100
## Iteration 50 / 100
## Iteration 60 / 100
## Iteration 70 / 100
## Iteration 80 / 100
## Iteration 90 / 100
## Iteration 100 / 100
preds <- boost_model$predict(X)
# Get variable importance
imp <- boost_model$variable_importance()
##
## Variable Importance
## ===================
## Feature Importance
## X1 53.8111025
## X2 44.6466990
## X3 0.9816176
## X4 0.5605810
## Feature Importance
## 1 X1 53.8111025
## 2 X2 44.6466990
## 3 X3 0.9816176
## 4 X4 0.5605810
# Plot loss history
boost_model$plot_loss()
# Cross-validation
cv_scores <- cross_val_score(boost_model, X, y, cv = 5)
## | | | 0% | |============== | 20% | |============================ | 40% | |========================================== | 60% | |======================================================== | 80% | |======================================================================| 100%