Classification II

Support vector machines

Published

October 2, 2026

Maximum margin classifiers; the soft-margin extension; kernel SVMs for non-linear decision boundaries. The lab goes further than the slides: interactive simulations of hyperplanes, support vectors, and kernels, and variable selection for classification with the lasso.

📖 Assigned reading: ISLR ch. 9

📊 Slides

🧪 Labs

Both labs run entirely in your browser, no local R install needed. Start with Lab 1.

  • 🎛️ Lab 1: Seeing SVMs (Shiny simulations) — sliders on simulated data: hyperplanes, support vectors, the cost dial, kernels, multi-class.
  • 🧪 Lab 2: Real tumor data (live R) — edit and re-run any code cell: SVMs on biopsy data, then variable selection with the lasso. A “Download my code” button captures whatever you’ve written before you leave.