Blake Mason

Blake Mason

Blake Mason is a Postdoctoral Research Fellow in Electrical and Computer Engineering at Rice University. Prior to coming to Rice, I was a PhD student at the University of Wisconsin–Madison in Electrical and Computer Engineering.

Publications

Nearly Optimal Algorithms for Level Set Estimation

Published in Artifical Intelligence and Statistics (AIStats), 2022

This paper provides the first instance-dependent, non-asymptotic upper bounds on sample complexity of level-set estimation that match information theoretic lower bounds.

Recommended citation: Mason, B., Camilleri, R., Mukherjee, S., Jamieson, K., Nowak, R., & Jain, L. (2021). Nearly Optimal Algorithms for Level Set Estimation. arXiv preprint arXiv:2111.01768. (To appear at AIStats, 2022) https://arxiv.org/pdf/2111.01768

Parameters or Privacy: A Provable Tradeoff Between Overparameterization and Membership Inference

Published in arXiv preprint arXiv:2202.01243, 2022

In this paper, we study an underexplored hidden cost of overparameterization: the fact that overparameterized models are more vulnerable to privacy attacks, in particular the membership inference attack that predicts the (potentially sensitive) examples used to train a model.

Recommended citation: Tan, J., Mason, B., Javadi, H., & Baraniuk, R. G. (2022). Parameters or Privacy: A Provable Tradeoff Between Overparameterization and Membership Inference. arXiv preprint arXiv:2202.01243. https://arxiv.org/pdf/2202.01243

Practical, Provably-Correct Interactive Learning in the Realizable Setting: The Power of True Believers

Published in Neural Information Processing Systems, 2021

We consider interactive learning in the realizable setting and develop a general framework to handle problems ranging from best arm identification to active classification.

Recommended citation: Katz-Samuels, J., Mason, B., Jamieson, K. G., & Nowak, R. (2021). Practical, Provably-Correct Interactive Learning in the Realizable Setting: The Power of True Believers. Advances in Neural Information Processing Systems, 34. https://proceedings.neurips.cc/paper/2021/file/94aada62f90dd50a84ca74304563d5db-Paper.pdf

Nearest neighbor search under uncertainty

Published in Uncertainty In Artificial Intelligence, 2021

This paper shows how ideas from cover trees and multi-armed bandits can be leveraged to develop an neariest neighbor search algorithm from noisy data that has optimal dependence on the dataset size and the (unknown) geometry of the dataset.

Recommended citation: Mason, B., Tripathy, A., & Nowak, R. (2021, December). Nearest neighbor search under uncertainty. In Uncertainty in Artificial Intelligence (pp. 1777-1786). PMLR. https://proceedings.mlr.press/v161/mason21a/mason21a.pdf

NFT-K: Non-Fungible Tangent Kernels

Published in arXiv preprint arXiv:2110.04945 - (To appear at ICASSP 2022), 2021

We develop a new network as a combination of multiple neural tangent kernels, one to model each layer of the deep neural network individually as opposed to past work which attempts to represent the entire network via a single neural tangent kernel.

Recommended citation: Alemohammad, S., Babaei, H., Barberan, C. J., Liu, N., Luzi, L., Mason, B., & Baraniuk, R. G. (2021). NFT-K: Non-Fungible Tangent Kernels. arXiv preprint arXiv:2110.04945. https://arxiv.org/pdf/2110.04945

Improved confidence bounds for the linear logistic model and applications to bandits

Published in International Conference on Machine Learning, 2021

In this paper, we propose improved fixed-design confidence bounds for the linear logistic model. With this bound we propose a new regret minimization and a new pure exploration bandit algorithm.

Recommended citation: Jun, K. S., Jain, L., Mason, B., & Nassif, H. (2021, July). Improved confidence bounds for the linear logistic model and applications to bandits. In International Conference on Machine Learning (pp. 5148-5157). PMLR. http://proceedings.mlr.press/v139/jun21a/jun21a.pdf

Learning nearest neighbor graphs from noisy distance samples

Published in Neural Information Processing Systems, 2019

We consider the problem of actively learning the nearest neighbor graph of a dataset of n items in as few queries as possible.

Recommended citation: Mason, B., Tripathy, A., & Nowak, R. (2019). Learning nearest neighbor graphs from noisy distance samples. Advances in Neural Information Processing Systems, 32. https://proceedings.neurips.cc/paper/2019/file/98c56bce74669e2e4e7a9fc1caa8c326-Paper.pdf

Cognitive Task Analysis for Implicit Knowledge About Visual Representations With Similarity Learning Methods

Published in Cognitive Science, 2019

This paper provides the first method to assess students’ perceptual competencies implicitly, without requiring verbalization or assuming explicit visual attention. These findings have implications for the design of instructional interventions that help students acquire perceptual representational competencies.

Recommended citation: Mason, B., Rau, M. A., & Nowak, R. (2019). Cognitive Task Analysis for Implicit Knowledge About Visual Representations With Similarity Learning Methods. Cognitive science, 43(9), e12744. https://onlinelibrary.wiley.com/doi/pdfdirect/10.1111/cogs.12744

Machine Beats Human at Sequencing Visuals for Perceptual-Fluency Practice.

Published in International Educational Data Mining Society, 2018

Here, we describe a novel educational data mining approach that uses machine learning to generate an optimal sequence of visuals for perceptual-fluency problems.

Recommended citation: Sen, A., Patel, P., Rau, M. A., Mason, B., Nowak, R., Rogers, T. T., & Zhu, X. (2018). Machine Beats Human at Sequencing Visuals for Perceptual-Fluency Practice. International Educational Data Mining Society. https://files.eric.ed.gov/fulltext/ED593113.pdf

How to Model Implicit Knowledge? Similarity Learning Methods to Assess Perceptions of Visual Representations

Published in International Educational Data Mining Society, 2016

In this paper, we develop a model of how undergraduate chemistry students perceive images of molecules.

Recommended citation: Rau, M. A., Mason, B., & Nowak, R. (2016). How to Model Implicit Knowledge? Similarity Learning Methods to Assess Perceptions of Visual Representations. International Educational Data Mining Society. https://files.eric.ed.gov/fulltext/ED592702.pdf

Thermoacoustic Transduction in Individual Suspended Carbon Nanotubes

Published in ACS Nano, 2015

We report an experimental measurement of the acoustic signal emitted from an individual suspended carbon nanotube (CNT) approximate 2 μm in length, 1 nm in diameter, and 10^–21 kg in mass.

Recommended citation: Mason, B. J., Chang, S. W., Chen, J., Cronin, S. B., & Bushmaker, A. W. (2015). Thermoacoustic transduction in individual suspended carbon nanotubes. ACS nano, 9(5), 5372-5376. https://pubs.acs.org/doi/full/10.1021/acsnano.5b01119