Matt Holland :: Info

I'm currently working as an Assistant Professor at Osaka University [ja], Institute of Scientific and Industrial Research [ja], in Osaka, Japan.

In a broad sense, my interests concentrate around the theoretical foundations of machine learning, and the ways in which this kind of very precise knowledge can be leveraged to design reliable statistical estimation procedures. That is, the pursuit of efficient algorithms that work for reasons we understand.


Publications  |  Contact  |  Misc.

For more detailed information, see my researchmap page.

Selected publications and related materials

Distribution-robust mean estimation via smoothed random perturbations.
Matthew J. Holland.
Preprint.
[arXiv, code, bib]

PAC-Bayes under potentially heavy tails.
Matthew J. Holland.
Preprint.
[arXiv, bib]

Better generalization with less data using robust gradient descent.
Matthew J. Holland and Kazushi Ikeda.
Presented at: ICML 2019. Proceedings: PMLR 97:2761-2770, 2019.
[PMLR, bib]

Robust gradient descent via back-propagation: A Chainer-based tutorial.
Matthew J. Holland.
Preprint.
[pdf, code]

Efficient learning with robust gradient descent.
Matthew J. Holland and Kazushi Ikeda.
Machine Learning (to appear).
[arXiv, doi, bib]

Robust descent using smoothed multiplicative noise.
Matthew J. Holland.
Presented at: AISTATS 2019. Proceedings: PMLR 89:703-711, 2019.
[arXiv, PMLR, bib]

Classification using margin pursuit.
Matthew J. Holland.
Presented at: AISTATS 2019. Proceedings: PMLR 89:712-720, 2019.
[arXiv, code, demo, PMLR, bib]

Robust regression using biased objectives.
Matthew J. Holland and Kazushi Ikeda.
Journal: Machine Learning, 106(9):1643-1679, 2017.
Oral: ECML-PKDD 2017, Skopje, Macedonia.

[pdf, code, doi, bib]

Minimum proper loss estimators for parametric models.
Matthew J. Holland and Kazushi Ikeda.
IEEE Transactions on Signal Processing, 64(3):704-713, 2016.
[pdf, data, doi, bib]

Location robust estimation of predictive Weibull parameters in short-term wind speed forecasting.
Matthew J. Holland and Kazushi Ikeda.
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2015.
Brisbane, Australia

[doi, bib]

Forecasting in wind energy applications with site-adaptive Weibull estimation.
Matthew J. Holland and Kazushi Ikeda.
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2014.
Florence, Italy

[doi, bib]

Contact

I can be reached at the following address:

Miscellany

In the past, I studied microeconomic theory and more practitioner-oriented finance. Having lived in Kyoto, Osaka and Nara for some time, I have also put a fair amount of effort into trying to achieve fluency in written and spoken Japanese. Now, besides research-related work, I spend most of my free time cycling, reading, or eating/drinking in Osaka.