Jiawei H
Research Scientist at Institute for Health Metrics and Evaluation
Beijing, United States
Summary
👤
Senior🎓
Top SchoolJiawei H is a research scientist with 9 years of experience applying deep learning, epidemiologic modeling, and big-data engineering to global health problems. Based in Beijing and working at IHME, Jiawei has led Gates Foundation-funded geospatial HIV mapping at 5×5 km resolution and hybridized mechanistic SEIR priors with transformer models to improve epidemic forecasts. He combines hands-on ML skills (LLM fine-tuning with PEFT and quantization, DeepSpeed, TabPFN, Chronos) with statistical rigor (Gaussian Markov Random Fields, boosted regression trees, Lasso/XGBoost) and has reengineered forecasting pipelines for 100× computational gains. Fluent in English, French, and Chinese, Jiawei translates complex research into policy-relevant insights and supervises cross-disciplinary teams. A less obvious strength is his knack for benchmarking foundational models from first principles—building GPT-2 from scratch and fine-tuning large open models—to both validate and accelerate applied public-health modeling.
9 years of coding experience
10 years of employment as a software developer
Doctor of Philosophy - PhD, Global Health Metrics, Doctor of Philosophy - PhD, Global Health Metrics at University of WashingtonBachelor's degree, Hydrology and Water Resources Science, Bachelor's degree, Hydrology and Water Resources Science at Wuhan University
English, Chinese, French
Github Skills (23)
partitioning8
3d-object-detection8
metric-learning8
re-identification8
image-retrieval8
video-understanding8
cross-domain8
pytorch7
point-cloud7
computer-vision6
deep-learning6
graph6
object-detection5
machine-learning5
object-tracking4
Programming languages (2)
Github contributions (5)
jiaweihe1996/jiaweihe1996.github.io
Sep 2020 - Sep 2025
Contributions:227 pushes, 7 branches in 5 years 1 month
jiaweihe1996/GMTracker
Jun 2021 - Dec 2021
Official PyTorch implementation of "Learnable Graph Matching: Incorporating Graph Partitioning with Deep Feature Learning for Multiple Object Tracking" (CVPR 2021).
Contributions:11 commits, 16 pushes, 8 branches in 6 months
graphobject-trackingpartitioningpytorchdeep-learning
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.