Summer Assessment Program
147 students · 86% completion
Scheduled and ran testing for 147 students across four home languages, then wrote the placement report used to set next year's cohorts.
147 students · 86% completion
Scheduled and ran testing for 147 students across four home languages, then wrote the placement report used to set next year's cohorts.
69% honest re-test · leak found and fixed
Built a model to predict WNBA winners, then re-tested it: found a data leak, rebuilt it on pre-game data only, and compared it with simple baselines.
140 years of Census data
Mapped demographic change in the U.S. from 1870 to 2010, including the Great Migration, using R.
10 models · 98% accuracy
Benchmarked ten models on the Wisconsin breast cancer dataset by precision, recall, F1 and ROC-AUC. Logistic regression reached 98% accuracy.
19,528 traffic samples
Trained a deep autoencoder on 19,528 network traffic samples to flag unusual traffic.
Making Waves Academy. Ran scheduling, family outreach and reporting for a 147-student assessment program.
Making Waves Academy. Built Google Sheets trackers and unit dashboards, managed compliance reporting for a $2.5M program budget, and presented data to school leadership.
Led a team of 4 through a 15-week redesign of a React.js course-planning app.
Built a Google Sheets add-on that replaced Excel for simulation work in math courses.
Google Sheets / Excel (FILTER, SORT, LET, IFERROR), Python (pandas, NumPy, scikit-learn), R (tidyverse, ggplot2), SQL, regression, Monte Carlo simulation
Program coordination, compliance reporting, roster and outreach tracking, deadline-driven work across departments, multilingual family communication
Dashboards, summary reports, presentations for leadership, Matplotlib, Seaborn, ggplot2
B.S. Applied Mathematics, UC Davis
M.A. Teaching, Secondary Mathematics, Western Governors University
melavinabeltran@gmail.com