Boshen Yang

Computer Science and Statistics student at UNC Chapel Hill

with a passion for AI, machine learning, computer vision, and data-driven solutions.

📧 yangbs@unc.edu

📱 (984) 325-4576

About Me

I'm a junior at the University of North Carolina at Chapel Hill pursuing a double major in Computer Science and Statistics & Operations Research. With a GPA of 3.93, I'm passionate about leveraging technology to solve complex problems, particularly in the areas of artificial intelligence, computer vision, and financial technology.

My interests span from developing AI-powered sports analytics systems to working on autonomous vehicle safety systems. I enjoy tackling challenging projects that combine theoretical knowledge with practical applications. When I'm not coding, I'm exploring the latest developments in machine learning and seeking opportunities to apply data science techniques to real-world problems.

Relevant Courses

COMP 426

Modern Web Programming

COMP 562

Machine Learning

COMP 301

Foundations of Programming

COMP 311

Computer Organization

COMP 421

Files and Databases

STOR 455

Methods of Data Analysis

STOR 320

Introduction to Data Science

STOR 435

Probability Theory

Technical Skills

Programming Languages

Java Python C C++ R SQL JavaScript HTML/CSS MATLAB

Tools & Frameworks

PyTorch React Node.js Git Docker MySQL MongoDB OpenCV YOLOv8

Featured Projects

Tactical-View AI: Soccer Analytics

Python • OpenCV • PyTorch • YOLOv8
  • Engineered an end-to-end AI pipeline in Python and PyTorch to analyze full-length soccer matches, automatically generating and exporting player-specific highlight clips.
  • Implemented a ResNet-18 scene classifier to automatically filter non-match footage(e.g., replays, audience), ensuring analysis was performed exclusively on relevant gameplay.
  • Fine-tuned a YOLOv8 object detection model on a specialized dataset to accurately identify players, referees, and the ball, overcoming the limitations of pre-trained models. Integrated ByteTrack to assign persistent IDs for robust multi-object tracking through occlusions.
  • Calculated real-world performance metrics by first compensating for camera movement with Optical Flow. Subsequently, implemented an automated perspective transformation using a YOLOv8 keypoint model to generate a Homography matrix, enabling the conversion of player positions to meter-scale coordinates for accurate speed and distance tracking, visualized in a final 2D radar view.
Computer Vision Machine Learning Sports Analytics

Autonomous Vehicle Safety System

MATLAB • Genetic Algorithm • SQP Optimization
  • Developed a kinematic model and obstacle avoidance system for a self-driving car-trailer to enable autonomous trajectory planning.
  • Designed and implemented a hybrid optimization strategy in MATLAB and iSight, combining a Genetic Algorithm (GA) with Successive Quadratic Programming (SQP) to significantly reduce computation time while ensuring path safety.
  • Co-authored and published the complete methodology and findings in the paper, "Tracking and Safety Control for Car-Trailer System with Optimization", presented at the ICHSSR 2022 conference.
Autonomous Systems Optimization Published Research

Professional Experience

Wealth Management Intern

CITIC Securities | Wuhan, China
Summer 2025
  • Researched and evaluated a diverse range of financial products and services, from mutual funds to sophisticated High-Net-Worth Individual solutions like family trusts and global investment platforms.
  • Conducted fundamental analysis on target equities using valuation models, including the Dividend Discount Model (DDM), to identify potential undervalued investment opportunities for client portfolios.
  • Collaborated within a team in the Wealth Management division to develop and present a final capstone project: a comprehensive asset allocation proposal for a model client, synthesizing insights from market analysis, product research, and simulated trading exercises.