Hands-on: k-NN Regression in Shiny

Overview

In this 30 min session you will:

  • Launch and explore the k-NN Shiny app

  • Adjust k, select backend (R vs C++)

  • Observe train/test performance and parity plots

Setup

Download the Shiny app:

curl -O https://raw.githubusercontent.com/mmadoliat/WSoRT/refs/heads/main/R/app.R

Launch the App

# Run the Shiny application
shiny::runApp('app.R')

Tasks

  1. Change the train/test split ratio in make_split() (e.g. 80/20).
  2. Add a slider input for the split ratio in ui and update server logic.
  3. Extend the performance panel to show Mean Absolute Error (MAE).