Efficient Machine Learning Model Query

We design methods that let you query compact ML models—Gaussian Processes (GPs) and Implicit Neural Representations (INRs)—without brute force. By using uncertainty and tight bounds, we skip work safely and return what you care about (iso-surfaces, statistics, parameter effects) much faster.

Why this matters

Modern simulations create massive data. We compress them into learned models to save space—but then waste compute on naïve queries (dense sampling, full parameter sweeps, heavy Monte Carlo).
This project improves the query side, making interrogation of ML models efficient, interactive, and reliable. (Li et al., 2024; Li & Shen, 2024; Chen et al., 2025)

References

2025

TVCG

Explorable inr: an implicit neural representation for ensemble simulation enabling efficient spatial and parameter exploration

Yi-Tang Chen, Haoyu Li, Neng Shi, and 3 more authors

IEEE Transactions on Visualization and Computer Graphics, 2025

Bib

@article{chen2025explorable,
  title = {Explorable inr: an implicit neural representation for ensemble simulation enabling efficient spatial and parameter exploration},
  author = {Chen, Yi-Tang and Li, Haoyu and Shi, Neng and Luo, Xihaier and Xu, Wei and Shen, Han-Wei},
  journal = {IEEE Transactions on Visualization and Computer Graphics},
  year = {2025},
  publisher = {IEEE},
}

2024

Efficient level-crossing probability calculation for Gaussian process modeled data

Haoyu Li, Isaac J Michaud, Ayan Biswas, and 1 more author

In 2024 IEEE 17th Pacific Visualization Conference (PacificVis), 2024

Bib

@inproceedings{li2024efficient,
  title = {Efficient level-crossing probability calculation for Gaussian process modeled data},
  author = {Li, Haoyu and Michaud, Isaac J and Biswas, Ayan and Shen, Han-Wei},
  booktitle = {2024 IEEE 17th Pacific Visualization Conference (PacificVis)},
  pages = {252--261},
  year = {2024},
  organization = {IEEE},
}

TVCG

Improving efficiency of iso-surface extraction on implicit neural representations using uncertainty propagation

Haoyu Li and Han-Wei Shen

IEEE Transactions on Visualization and Computer Graphics, 2024

Bib

@article{li2024improving,
  title = {Improving efficiency of iso-surface extraction on implicit neural representations using uncertainty propagation},
  author = {Li, Haoyu and Shen, Han-Wei},
  journal = {IEEE Transactions on Visualization and Computer Graphics},
  volume = {31},
  number = {2},
  pages = {1513--1525},
  year = {2024},
  publisher = {IEEE},
}