Gain insights into essential machine learning algorithms in R, including CART, random forest, and XGBoost. Discover model tuning and cross-validation to create accurate, robust data science models.

Dockerfile.tar.gz

tidymodels

The R programming language is widely used in the field of data science. Machine learning is a fundamental skill for learners looking to master industry algorithms in the field of data science.

In this course, you’ll learn about several essential algorithms used in machine learning, including classification and regression trees (CART), random forest, and XGBoost. CART is a decision tree algorithm that’s used for both classification and regression problems. Random forest is an ensemble learning method that uses multiple decision trees to improve the accuracy of predictions. XGBoost, short for Extreme Gradient Boosting, is a powerful algorithm that’s also used for regression and classification problems. You’ll also learn about cross-validation and model tuning, which are essential skills for crafting valuable machine learning models. 

After taking this course, you’ll have the crucial skills to ensure that the machine learning models you create are accurate, robust, and reliable.

Data Science with R: Decision Trees and Random Forests

Supported xgboost Hyperparameters

align-center

#================================================================================================
# Specifying gradient boosting for classification
#
titanic_model <- boost_tree() %>%
  set_engine("xgboost", event_level = "first", verbose = 1) %>%
  set_mode("classification")

titanic_model

Learn how to tune the XGBoost algorithm using tidymodels.

Hyperparameter	Description	Default Value
`mtry`	The number of randomly selected predictive features used	The count of predictive features
`trees`	The number of trees used in the ensemble	15
`min_n`	The minimum number of observations required to make a split	1
`tree_depth`	The maximum depth (i.e., number of splits) of ensemble trees	6
`learn_rate`	The weight applied to each ensemble tree's predictions	0.3
`loss_reduction`	The minimum amount by which a new tree must improve the ensemble for the algorithm to continue	0.0
`sample_size`	The proportion of data sampled to train each ensemble tree (a value of 1.0 denotes 100 percent of the data)	1.0
`stop_iter`	The number of iterations before stopping	Infinite

Welcome to the Course

Supervised Learning

Classification Tree Math

Using Classification Trees in R

Introducing the Bias-Variance Tradeoff

Model Tuning

Model Tuning with tidymodels

Feature Engineering

Regression Trees

The Random Forest Algorithm

Using Random Forests

Gradient Boosting Trees

Continuing Your Journey

Credit Card Fraud Detection using the R Language

Tuning XGBoost with tidymodels

XGBoost hyperparameters

Supported xgboost Hyperparameters