University of California, Riverside
Focused on embedded systems, machine learning, robotics, and systems-oriented engineering work across Python, C/C++, Linux, computer vision, and hardware integration.
Tae Jong Kang
UC Riverside Electrical Engineering
I am working on prediction market algo trading
Featured Analysis
More on the Blog
Monte Carlo · Betting systems
What half-Kelly does to a $1,000 tennis bankroll
1,000 simulated trading days on an AUC 0.73 ATP model — a small, real edge buried under heavy variance, and why sizing matters more than selection.
Read the analysisProjects
View AllXGBoost Tennis Model
XGBoost win-probability model with a live trading layer, evaluated honestly on real settled bets against prediction-market prices.
Computer Vision Coffee Sorting
Raspberry Pi and OpenCV system for visual sorting, sensor feedback, and real-time hardware control.
CUDA Mixture of Experts
GPU-accelerated tennis Monte Carlo simulator using CUDA C++, Mixture of Experts routing, softmax expert blending, and CPU vs GPU benchmark comparison.
Professional Experience
View AllIndependent Technical Projects
Designed complete workflows from data ingestion and modeling through interface design and deployment structure.
Blog
View AllA Monte Carlo study of a half-Kelly staking plan on an AUC 0.73 ATP match model: bankroll growth, intraday drawdown, and how full, half, and quarter Kelly compare over a 15-bet day.
Serve-based tennis modeling paper using Wsp, dynamic Monte Carlo simulation, synthetic training data, and machine-learning regression to estimate total games, win probability, and over/under outcomes.
Ethical, food-safety, cybersecurity, economic, environmental, and societal impacts of a low-cost optical coffee bean sorter with robotic arm.
How rapid prototyping with PLA/FDM helped build the sorter, and why commercial food-contact designs require safer materials and manufacturing methods.