Archive
For reference.
Past projects, blog posts, and talks. Held aside from the main listings but kept reachable.
Projects
Data viz
Olympic 100m Dashboard
A Quarto dashboard for the men's 100m — built for Posit::conf 2024 to show that pure Quarto (no Shiny) is enough for polished analytics.
ML
Symptom Severity Classification
Python Random Forest mapping symptom flags to mild / moderate / severe ratings, with extracted rules for clinician review.
ML
Predicting $50K+ Salaries
Team ML project on the OpenML census data — PCA feature selection plus a tuned Random Forest reaching kappa 0.95.
ML
Customer Return Forecasting
Feature-engineered Random Forest that lifted return prediction from AUC 0.625 to 0.98 on retail data.
ML
Pinot Province Prediction
NLP-flavored Random Forest that infers a Pinot Noir's province of origin from a critic's tasting notes.
Statistics
U.S. Healthcare Spending, 1980–2014
Longitudinal analysis of U.S. healthcare expenditure by category, region, and state across 35 years.
Blog
Data viz
Apple's Journey in the Stock Market
Interactive Plotly walkthrough of Apple stock from 1981 to present, with custom theming.
Data viz
Do You Like Stretching? I Would Reconsider
Plotly exploration of a large gym dataset — what users actually rate highest, and what they do not.
Data viz
Rating Pinot Wines: Is More Expensive Better?
Advanced tabular display in R with DT, gt, and gtExtras — using Pinot Noir as the case study.
Data engineering
Pokedex Database
A two-week data-engineering sprint to build a normalized PostgreSQL Pokedex from Kanto, Johto, and Hoenn data.
Dashboard
College Debt Dashboard
A four-tool Shiny app for prospective students: salary estimator, tuition estimator, debt estimator, and debt calculator.
Statistics
The Real Cost of Smoking: Health Insurance Premiums
Quantifying how smoking and BMI drive up insurance premiums by age and gender, in Python.
Talks
No archived talks yet.