When Predictors Outnumber Data: Making Sense of High-Dimensional Regression

Published:

Built an interactive Plotly Dash blog site that pairs modular Markdown notes with live demos to unpack high-dimensional regression, starting from rank diagnostics to LASSO paths and economic intuition. My groupmates and I created it to help classmates grasp why classical regression breaks when features explode, and how modern regularization tools step in.

Highlights

  • Live Markdown sections sourced from notes/ and composed via the Dash app entry point (main.py).
  • Reusable components in components/ for visualizing LASSO selection, full-rank checks, and regression workflows.
  • Cohesive styling through assets/styles.css and shared design tokens in theme.py.

Explore

This was a class project for MSDS 601 Linear Regression.