Finance Certifications Goldman Sachs and Amazon Teams Trust
AI, Data Science & Cloud Certificates from Google, IBM & Meta
Overview
Google, IBM & Meta Certificates – 40% Off
One Coursera Plus subscription covers most Professional Certificates on Coursera.
Unlock All Certificates
This advanced screencast demonstrates custom feature engineering, hyperparameter tuning, and model explainability for an XGBoost regression model predicting board-game ratings. It uses R and tidy modeling tools with SHAP-based feature-importance and partial-dependence plots.
Syllabus
Introduction
Data overview
Average distribution
Modeling
Preprocessing
Custom Tokenizing
String Squish
Regression
Tuning
Results
Plotting function game
Finding the best game
Last fit
Testing set
Explainability tools
parsnip fit
model importance
shop
model
other arguments
matrix
making plots
dependency partial plot
min age plot
summary
Taught by
Julia Silge