Formula-first walkthroughs for core analytics modeling ideas. Each page pairs the modeling question, the math, the R code, and the judgment calls that matter when the method meets real data.
This is an independent educational resource. It is designed for learners who want to understand how analytics models work rather than only call a library.
- Classification: SVM and KNN
- Validation: train/test splits and cross-validation
- Clustering: k-means
- Change detection: CUSUM
- Time series: exponential smoothing, ARIMA, and GARCH
- Regression: linear and logistic regression
- Transformations: PCA and Box-Cox
- Trees: CART
- Variable selection, design of experiments, simulation, missing data, optimization, and selected advanced topics
Install Quarto and R, then restore package dependencies:
Rscript -e "renv::restore()"Preview the site:
quarto previewRender the static site:
quarto renderRendered output is written to _site/, which is intentionally ignored by Git.
This repo is configured for GitHub Pages through GitHub Actions. On each push
to main, the workflow restores the R environment with renv, renders the
Quarto site, uploads _site/, and deploys it to Pages.
After creating the GitHub repository, configure Pages to use GitHub Actions as the source.
Expected project-site URL:
https://<github-user>.github.io/analytics-modeling/
_quarto.yml- Quarto website configuration*.qmd- source pagesstyles.css- site styling.github/workflows/publish.yml- GitHub Pages deployment workflowrenv.lock- reproducible R package environment
Generated folders such as _site/, .quarto/, and scratch-files/ are not
part of the public source repository.
This repository uses a dual-license model:
- Educational prose, explanations, tables, and generated figures are licensed under Creative Commons Attribution-NonCommercial 4.0 International.
- Code, configuration, CSS, and workflow files are licensed under MIT.
See LICENSE.md for details.