synthetic_ensemble_ML.R

library(mlr)

## Synthetic data ----
set.seed(200)
n = 100
x <- 1:n
y <- x + rnorm(n = 50, mean = 15, sd = 15)
dat <- data.frame(x,y)
## newdata:
newdata <- data.frame(
  x = -100:200
)
## Fir Ensemble ML using mlr ----
SL.library = c("regr.ranger", "regr.glm", "regr.gamboost", "regr.ksvm")
lrns <- lapply(SL.library, mlr::makeLearner)
tsk <- mlr::makeRegrTask(data = dat, target = "y")
init.m <- mlr::makeStackedLearner(lrns, method = "stack.cv", super.learner = "regr.lm", resampling=mlr::makeResampleDesc(method = "CV"))
eml = train(init.m, tsk)
summary(eml$learner.model$super.model$learner.model)
# Residuals:
#   Min       1Q   Median       3Q      Max
# -25.8665  -7.1403   0.0125   6.9210  27.7502
#
# Coefficients:
#   Estimate Std. Error t value Pr(>|t|)
# (Intercept)    -2.3088     2.8341  -0.815   0.4173
# regr.ranger    -0.4184     0.2201  -1.901   0.0603 .
# regr.glm        4.7697     0.9230   5.168 1.31e-06 ***
# regr.gamboost  -4.1556     1.0181  -4.082 9.31e-05 ***
# regr.ksvm       0.8295     0.3419   2.426   0.0171 *
#   ---
# Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
#
# Residual standard error: 10.89 on 95 degrees of freedom
# Multiple R-squared:  0.8796,	Adjusted R-squared:  0.8745
# F-statistic: 173.5 on 4 and 95 DF,  p-value: < 2.2e-16
## GLM is the best model, RF and kSVM are on the edge