Mingyuan Zhang
Teaching Assitant at National University of Singapore, Faculty of Science
Singapore, Singapore
Summary
🤩
Rockstar🎓
Top SchoolMingyuan Zhang is a third-year Data Science and Analytics student at NUS with four years of hands-on experience applying statistical methodology, operations research, and machine learning to real problems. As a teaching assistant, he has guided over 120 students in R-based data manipulation and visualization and taught foundational ML concepts, demonstrating strong communication and pedagogical skills. As a research assistant he developed a Python package for hierarchical LDA and applied hLDA with covariates to uncover community structure in network data, highlighting an ability to turn statistical theory into reusable tooling. He also contributes to open-source ML work, enhancing the well-known mmhuman3d repository with loss functions and training components for 3D human model estimation. Proficient in Python, R, SQL, and Java and familiar with Scikit-learn, TensorFlow, and Tidyverse, he focuses on robust model building, simulation, and statistical inference for practical impact. Mingyuan seeks collaborative opportunities in data science, statistics, and finance where methodological rigor meets applied outcomes.
5 years of coding experience
Bachelor of Science - BS, Data Science and Analytics, Bachelor of Science - BS, Data Science and Analytics at National University of Singapore
Github Skills (9)
computer-vision10
pytorch10
machine-learning10
loss-functions10
trainings10
hmr10
python10
modeling10
resnet9
Programming languages (1)
Github contributions (5)
open-mmlab/mmhuman3d
Aug 2021 - Jul 2022
OpenMMLab 3D Human Parametric Model Toolbox and Benchmark
Role in this project:
ML Engineer
ML EngineerContributions:17 reviews, 22 commits, 16 PRs in 11 months
Contributions summary:Mingyuan primarily contributed to the `mmhuman3d` repository by adding and modifying code related to 3D human model estimation, especially with HMR and SPIN models. They added loss functions (SmoothL1Loss, MSELoss) and implemented components for model training. The user also refactored code, renamed files, and integrated new model architectures, enhancing the functionality of the project.
3dopenmmlab
MotrixLab/MotionDiffuse
Sep 2022 - Nov 2022
MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model
Contributions:14 commits, 13 pushes, 1 branch in 2 months
3d-generationdiffusion-modelmotion-generationtext-driven
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.