Toby Dylan Hocking, PhD
Statistical machine learning researcher, focusing on fast optimization algorithms for big data

toby.hocking@nau.edu
(928)523-5209
School of Informatics, Computing, and Cyber Systems
Northern Arizona University
Building 90, Room 210
1295 S. Knoles Dr.
Flagstaff, AZ 86011
Jobs
My lab is recruiting masters and PhD students who are interested in working on new statistical models, optimization algorithms, interactive systems, and software for machine learning. If you are interested in joining, please first read my mentorship plan to see what I will expect of you, then read the application instructions.
Brief CV
- 2018-present: Assistant Professor, Northern Arizona University, School of Informatics, Computing, and Cyber Systems.
- 2014-2018: Postdoc, McGill University, Human Genetics Department, with Guillaume Bourque.
- 2013: Postdoc, Tokyo Institute of Technology, Computer Science Department, with Masashi Sugiyama.
- 2009-2012: PhD in Mathematics from Ecole Normale Supérieure de Cachan, with Francis Bach and Jean-Philippe Vert.
- 2008-2009: Masters student, Université Paris 6, Statistics Department, research internship at INRA Jouy-en-Josas with Mathieu Gautier and Jean-Louis Foulley.
- 2006-2008: research assistant at Sangamo BioSciences.
- 2002-2006: undergraduate, UC Berkeley, Double major in Molecular & Cell Biology and Statistics, honors thesis with Terry Speed.
- Full CV, Short Bio.
Research interests: fast and accurate algorithms for convex optimization (clustering, regression, ranking, classification) and discrete optimization (changepoint detection, dynamic programming). The main application domain for the algorithms I develop are in genomic data analysis (DNA copy number, ChIP-seq, etc); other applications include neuroscience, audio, internet, sensors, recommendation and ranking systems.
I think reproducible research is important, so in addition to every paper I write, I also provide
- A reference implementation of the algorithm(s) described in the paper, typically as an R package.
- Source code for doing the analyses and creating the figures, typically in a GitHub repo.
For more info, see my publications and software.
If you want to send me encrypted/signed messages, you can use my GPG key (fingerprint 1D46 6295 2738 32E6 F70B 9F64 45B0 8611 CDB1 FA96).
My ORCID is 0000-0002-3146-0865.
news
Aug 18, 2022 | Our textbook “Land Carbon Cycle Modeling: Matrix Approach, Data Assimilation, and Ecological Forecasting” was published by Taylor and Francis, my related materials. |
Jul 7, 2022 | Our paper about Labeled Optimal Partitioning has been published in Computational Statistics, Video. |
May 23, 2022 | Our paper about Survival Regression with Accelerated Failure Time Model in XGBoost has been published in Journal of Computational and Graphical Statistics, Documentation, Video. |
May 9, 2022 | I will be at the IMS meeting in London, chairing session TC8: New changepoint algorithms and theoretical results (Mon 27 June 2022, 16:00-17:30) and presenting in session TC9: New models and evaluation methods for changepoint detection (Tues 28 June 2022, 16:00-17:30). My talk title: Optimizing ROC Curves with a Sort-Based Surrogate Loss for Binary Classification and Changepoint Detection, slides: PDF. |
Feb 7, 2022 | Our paper about Generalized Functional Pruning Optimal Partitioning (GFPOP) for Constrained Changepoint Detection in Genomic Data was published in Journal of Statistical Software, Video. |