4
勾配ブースティングマシンの精度は、反復回数が増えると低下します
caretR のパッケージを介して勾配ブースティングマシンアルゴリズムを試しています。 小さな大学入学データセットを使用して、次のコードを実行しました。 library(caret) ### Load admissions dataset. ### mydata <- read.csv("http://www.ats.ucla.edu/stat/data/binary.csv") ### Create yes/no levels for admission. ### mydata$admit_factor[mydata$admit==0] <- "no" mydata$admit_factor[mydata$admit==1] <- "yes" ### Gradient boosting machine algorithm. ### set.seed(123) fitControl <- trainControl(method = 'cv', number = 5, summaryFunction=defaultSummary) grid <- expand.grid(n.trees = seq(5000,1000000,5000), interaction.depth = 2, shrinkage …
15
machine-learning
caret
boosting
gbm
hypothesis-testing
t-test
panel-data
psychometrics
intraclass-correlation
generalized-linear-model
categorical-data
binomial
model
intercept
causality
cross-correlation
distributions
ranks
p-value
z-test
sign-test
time-series
references
terminology
cross-correlation
definition
probability
distributions
beta-distribution
inverse-gamma
missing-data
paired-comparisons
paired-data
clustered-standard-errors
cluster-sample
time-series
arima
logistic
binary-data
odds-ratio
medicine
hypothesis-testing
wilcoxon-mann-whitney
unsupervised-learning
hierarchical-clustering
neural-networks
train
clustering
k-means
regression
ordinal-data
change-scores
machine-learning
experiment-design
roc
precision-recall
auc
stata
multilevel-analysis
regression
fitting
nonlinear
jmp
r
data-visualization
gam
gamm4
r
lme4-nlme
many-categories
regression
causality
instrumental-variables
endogeneity
controlling-for-a-variable