Chao Zhang
Georgia Institute of Technology, GA·Computer Science·Early-career
Many systems that matter are networks spread across space and changing over time, such as traffic grids, power grids, mobility patterns that carry disease, and molecules. Deep learning can forecast how these systems behave.
- Fields
Computer Science
Mathematics & Statistics
Engineering
- Topics
machine learning
deep learning
uncertainty quantification
- Grants
- 2 known grants, through Apr 2027
What they work on
- Develop deep predictive models for spatial networks that capture how they change over space and time while estimating the uncertainty of their own forecasts.
- Integrate these uncertainty-aware predictors into optimization and generation algorithms to search large design spaces for spatial networks, such as traffic, vaccine-distribution and molecular networks.
- Use model uncertainty to choose which data to collect next, so that spatial network design models can learn well from few labeled examples.
- Combine deep sequence models such as recurrent networks and transformers with Gaussian processes, a statistical method with built-in error estimates, to quantify uncertainty in time series predictions.
- Learn calibrated time series models and separate their sources of uncertainty, then use uncertainty to improve the reliability of public health forecasting and mobile-sensing analysis.
More about the research
However, it often cannot tell when its own predictions are unreliable, and it struggles to guide design decisions when data are scarce. Two projects tackle this together. They build machine learning models that report how uncertain they are, and they use that uncertainty to make better forecasts and better designs.
Why it matters
Forecasting models that cannot recognize their own errors can lead to bad decisions in healthcare and other high-stakes settings. Models that are aware of their uncertainty could make public health forecasting and mobile-sensing analysis more trustworthy. They could also help planners design traffic networks that reduce congestion, target vaccine distribution to contain outbreaks, and find molecules for greener materials.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Yao Zheng
University of Connecticut, CT·Mathematics & Statistics·Early-career
Develop recurrent neural network (RNN) models for high-dimensional time series that stay interpretable through Granger causality, a test of whether one series predicts another, to balance accuracy and interpretability.
Qi Yu
University of California-San Diego, CA·Computer Science·Early-career
Develop deep learning models that forecast spatiotemporal dynamics while obeying physical laws, with the goal of simulating ocean currents, traffic and epidemic spread faster than numerical solvers.
Minge Xie
Rutgers University-New Brunswick, NJ·Mathematics & Statistics
Develop PCSI (Predictability-Computability-Stability Inference) for binary and multi-class classification, screening many algorithms by predictive performance and using bootstrap resampling to improve uncertainty estimates, robustness and accuracy.
The school
Georgia Institute of Technology: 50 labs on PI Pages, $1.5B a year on research, 572 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.