David Hong

About

I am currently an NSF Postdoctoral Fellow in the Department of Statistics and Data Science at the Wharton School of the University of Pennsylvania, hosted by Edgar Dobriban. Previously, I was an NSF Graduate Research Fellow in the Department of Electrical Engineering and Computer Science at the University of Michigan, where I completed my PhD ("Learning Low-Dimensional Models for Heterogeneous Data") under the supervision of Jeff Fessler and Laura Balzano. I also spent a summer as a Data Science Graduate Intern at Sandia National Labs, where I worked with Tammy Kolda and Cliff Anderson-Bergman.

My current research interests include:

• Low-rank matrix and tensor estimation
• Heterogeneous, messy, and mis_ing data
• Large-dimensional and multivariate statistics
• Random matrix theory
• Statistical machine learning
• Imaging and inverse problems

Education

• Ph.D. in Electrical Engineering: Systems, University of Michigan, 2019.
• M.S. in Electrical Engineering: Systems, University of Michigan, 2015.
• B.S. in Electrical and Computer Engineering, Mathematics, Duke University, 2013.

Software

• HePPCAT.jl: Julia package for computing probabilistic PCA for samples with heterogeneous quality.
• WeightedPCA.jl: Julia package for performing optimally weighted PCA for samples with heterogeneous quality.
• FlipPA.jl: Julia package for estimating the rank of heterogeneous, noisy data using parallel analysis with signflips.
• GCPDecompositions.jl: Julia package for computing generalized CP tensor decompositions.
• ConvolutionalOperatorLearning.jl: Julia package for convolutional analysis operator learning.

Preprints

(* = equal contributors)

1. Selecting the number of components in PCA via random signflips
   David Hong*, Yue Sheng*, Edgar Dobriban.
   Under review.
   arXiv code software

Refereed Journal Papers

(* = equal contributors)

1. Optimally Weighted PCA for High-Dimensional Heteroscedastic Data
   David Hong, Fan Yang, Jeffrey A. Fessler, Laura Balzano.
   To appear in the SIAM Journal on Mathematics of Data Science, 2022.
   arXiv code software
2. Provable tradeoffs in adversarially robust classification
   Edgar Dobriban*, Hamed Hassani*, David Hong*, Alexander Robey*.
   To appear in the IEEE Transactions on Information Theory, 2022.
   arXiv
3. Group testing via hypergraph factorization applied to COVID-19
   David Hong, Rounak Dey, Xihong Lin, Brian Cleary, Edgar Dobriban.
   Nature Communications, 2022.
   code doi medRxiv talk web app
4. HePPCAT: Probabilistic PCA for Data with Heteroscedastic Noise
   David Hong, Kyle Gilman, Laura Balzano, Jeffrey A. Fessler.
   IEEE Transactions on Signal Processing, 2021.
   arXiv code doi software
5. Using viral load and epidemic dynamics to optimize pooled testing in resource-constrained settings
   Brian Cleary*, James A. Hay*, Brendan Blumenstiel, Maegan Harden, Michelle Cipicchio, Jon Bezney, Brooke Simonton, David Hong, Madikay Senghore, Abdul K. Sesay, Stacey Gabriel, Aviv Regev, Michael J. Mina.
   Science Translational Medicine, 2021.
   doi medRxiv
6. Subspace clustering using ensembles of K-subspaces
   John Lipor*, David Hong*, Yan Shuo Tan, Laura Balzano.
   Information and Inference, 2021.
   arXiv doi
7. Baseline estimation of commercial building HVAC fan power using tensor completion
   Shunbo Lei, David Hong, Johanna L. Mathieu, Ian A. Hiskens.
   Electric Power Systems Research, 2020.
   arXiv doi
8. Stochastic Gradients for Large-Scale Tensor Decomposition
   Tamara G. Kolda, David Hong.
   SIAM Journal on Mathematics of Data Science, 2020.
   arXiv doi
9. Generalized Canonical Polyadic Tensor Decomposition
   David Hong, Tamara G. Kolda, Jed A. Duersch.
   SIAM Review, 2020.
   arXiv doi software
10. Convolutional Analysis Operator Learning: Dependence on Training Data
    Il Yong Chun*, David Hong*, Ben Adcock, Jeffrey A. Fessler.
    IEEE Signal Processing Letters, 2019.
    arXiv doi
11. Asymptotic performance of PCA for high-dimensional heteroscedastic data
    David Hong, Laura Balzano, Jeffrey A. Fessler.
    Journal of Multivariate Analysis, 2018.
    arXiv doi
12. Closed-Form Expressions for Minimizing Total Harmonic Distortion in Three-Phase Multilevel Converters
    David Hong, Sanzhong Bai, Srdjan M. Lukic.
    IEEE Transactions on Power Electronics, 2013.
    doi

Refereed Conference Papers in Conference Proceedings

1. Generic Properties of Koopman Eigenfunctions for Stable Fixed Points and Periodic Orbits
   Matthew D. Kvalheim, David Hong, Shai Revzen.
   24th International Symposium on Mathematical Theory of Networks and Systems (MTNS), 2020.
   arXiv doi
2. Probabilistic PCA for Heteroscedastic Data
   David Hong, Laura Balzano, Jeffrey A. Fessler.
   2019 IEEE 8th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP), 2019.
   doi software
3. Incorporating Handcrafted Filters in Convolutional Analysis Operator Learning for Ill-Posed Inverse Problems
   Caroline Crockett, David Hong, Il Yong Chun, Jeffrey A. Fessler.
   2019 IEEE 8th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP), 2019.
   doi software
4. Exploration of tensor decomposition applied to commercial building baseline estimation
   David Hong, Shunbo Lei, Johanna L. Mathieu, Laura Balzano.
   2019 IEEE Global Conference on Signal and Information Processing (GlobalSIP), 2019.
   doi
5. Learning Dictionary-Based Unions of Subspaces for Image Denoising
   David Hong, Robert P. Malinas, Jeffrey A. Fessler, Laura Balzano.
   2018 26th European Signal Processing Conference (EUSIPCO), 2018.
   doi
6. Online Estimation of Coherent Subspaces with Adaptive Sampling
   Greg Ongie, David Hong, Dejiao Zhang, Laura Balzano.
   2018 IEEE Statistical Signal Processing Workshop (SSP), 2018.
   doi
7. Enhanced online subspace estimation via adaptive sensing
   Greg Ongie, David Hong, Dejiao Zhang, Laura Balzano.
   2017 51st Asilomar Conference on Signals, Systems, and Computers, 2017.
   doi
8. Towards a theoretical analysis of PCA for heteroscedastic data
   David Hong, Laura Balzano, Jeffrey A. Fessler.
   2016 54th Annual Allerton Conference on Communication, Control, and Computing (Allerton), 2016.
   arXiv doi