Training Sets, Test Sets, R, and ggplot

Evaluating Regression Model Performance in R

This course focuses on a specific, practical skill: splitting data into training and test sets to properly evaluate a regression model's performance, using R and ggplot throughout. Beyond the statistics, it builds general R programming ability along the way, including writing functions and for loops.

What You'll Practice

• Row-slicing data frames to create training and test sets
• Using R's predict function to generate and evaluate model predictions
• Adding titles and labels to ggplot visualizations
• Writing for loops and functions of two variables in R

Who It's For

Learners who already have a background in R, ggplot, and regression — ideally from prior linear and polynomial regression coursework — and want to build toward proper model evaluation. At a relaxed pace, it's designed to take about two weeks.

Key takeaway: a regression model that fits your training data well isn't automatically a good model — properly separating training and test sets is what tells you whether it actually generalizes.