Adam Zhu
Adam Zhu is an undergraduate at Carnegie Mellon University studying Statistics
and Machine Learning. His work includes usage-analytics research at NIST,
TrueLine (an iOS computer-vision bowling tracker), forensic image research at
CSAFE, statistical analysis of VR evacuation studies, and a co-authored
publication on matched binary diagnostic data.
Selected work
-
NIST, Research Data and Computing Office (SURF research intern, summer 2026):
built a reproducible five-stage Python pipeline (pandas, scikit-learn)
recovering reliable usage metrics for NIST's public data portal: per-IP
behavioral feature engineering and four classifiers (best F1 = 0.95). Found
51% of 190,687 logged requests were automated, exposed a disguised scraper
behind 50,000+ spoofed requests via a second-pass behavioral detection rule,
and showed 23% of 20,790 human searches returned no results.
-
TrueLine (Jan 2026–present): iOS app that turns one iPhone into a bowling ball
tracker: per-throw launch speed, board at the arrows, breakpoint, and entry
angle. Fine-tuned YOLOv8 exported to Core ML, Kalman-filter tracking with
Savitzky–Golay smoothing, four-corner calibration homography; board position
verified within ~1 board and speed within 1–2% against the Python/OpenCV
prototype. Swift, SwiftUI, Core ML, AVFoundation, PyTorch.
GitHub repo. Released on the
App Store in August 2026.
-
CSAFE, Iowa State: forensic analysis on 24,000+ camera images for device
fingerprinting and handwriting similarity datasets; co-author on a manuscript
under review at Forensic Sciences.
-
VR Applications Center (VRAC), Iowa State: regression models of evacuation
time across 454 trials and 227 participants, testing automated communication
strategies in simulated school-shooting evacuations; co-author on a manuscript
in preparation.
-
COR Robotics: co-designed curriculum and taught hands-on robotics (drones,
engineering) to 100+ students in grades 3–8.
-
Team Neutrino #3928 (FIRST): co-captain coordinating seven sub-teams and 40+
members; multiple NASA Engineering Inspiration Awards and Worlds
qualifications.
-
Publication: Lin, H., Zhu, A., Wang, C. (2024). "Statistical Tests for
Proportion Difference in One-to-Two Matched Binary Diagnostic Data:
Application to Environmental Testing of Salmonella in the United States."
Mathematics 12(5), 741.
doi.org/10.3390/math12050741.
Contact