Explore how to use tidymodels to tune a decision tree model for predicting shortcut usage in Mario Kart world records, and learn to analyze partial dependence profiles using DALEX. Walk through the process of data exploration, model setup, metric selection, prediction generation, and tree optimization. Discover techniques for customizing explainers and gain insights into world record times. Follow along with the step-by-step demonstration, from introduction to conclusion, and access the accompanying code on Julia Silge's blog for a comprehensive understanding of the analysis.
Overview
Syllabus
Introduction
Data
Exploratory plot
Data setup
Metrics
Predictions
Choose tree
Lastfit
Explainer
Customization
Conclusion
Taught by
Julia Silge
Related Courses
-
Model Canadian Wind Turbine Capacity With Decision Trees and Tidymodels
-
Modeling Hotel Bookings in R Using Tidymodels and Recipes
-
Partial Dependence Plots in Machine Learning Explainability - Day 17 of Kaggle's 30 Days of ML
-
Tune and Evaluate a Multiclass Lasso Model for NBER Working Papers
-
Predict Housing Prices in Austin TX with Tidymodels and XGBoost
-
Build a Predictive Text Model for Avatar: The Last Airbender with Tidymodels
