Abhyuday Mandal
University of Georgia, GA·Mathematics & Statistics
Some experiments depend on both how much of each ingredient is used and the order in which ingredients are added. In chemotherapy, for example, researchers must tune both the doses of several drugs and the sequence in which they are given.
- Fields
Mathematics & Statistics
Computer Science
Engineering
- Topics
design of experiments
quantitative-sequence experiments
order-of-addition experiments
- Grants
- 1 known grant, through Jul 2027
What they work on
- Develop Gaussian process models for physical experiments that combine quantitative factors like drug doses with sequence factors like drug order, as a basis for prediction and uncertainty estimates.
- Create active learning procedures that adaptively choose the next experimental run using the new Gaussian process models, to find optimal QS settings with fewer runs.
- Establish a Hopfield process framework to interpolate binary and categorical outputs of computer experiments, such as metal additive manufacturing success or failure, with uncertainty quantification.
- Construct optimal QS experimental designs by combining Williams-transformed good lattice point sets that are space-filling, orthogonal and paired-balanced.
More about the research
These quantitative-sequence (QS) experiments are hard to plan because the number of possible orders grows exponentially as drugs are added. With limited budgets, researchers need statistical tools to find the best settings with few runs. This collaborative project aims to build those tools for design, modeling, uncertainty quantification and active learning.
Why it matters
Experiments that tune both amounts and orders can be costly. Examples include finding chemotherapy drug combinations and setting up metal additive manufacturing processes. More efficient designs and adaptive methods could reduce the number of runs these experiments need. That could save resources for cancer researchers and for manufacturers.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Steven Kurtz
Drexel University, PA·Biomedical Engineering
Develop and optimize a hybrid additive manufacturing process for carbon-fiber-reinforced polyaryletherketones, combining 3D printing with heat-and-pressure consolidation, using design of experiments.
Mengdi Wang
Princeton University, NJ·Computer Science
Develop distributionally robust training methods for model-based reinforcement learning so that policies avoid unsafe states even when the real environment differs from the training data.
Chao Zhang
Georgia Institute of Technology, GA·Computer Science·Early-career
Develop deep predictive models for spatial networks that capture how they change over space and time while estimating the uncertainty of their own forecasts.
The school
University of Georgia: 27 labs on PI Pages, $628M a year on research, 598 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.