Wesley Tansey

tansey@cs.utexas.edu

academic cv

github

hacker news

stack exchange

twitter


Update July 2017! I am now a postdoc at Columbia! I am officially in Raul Rabadan's lab in Systems Biology; I also work with Chris Wiggins and David Blei at the Data Science Institute. My work at Columbia is focused on statistical methods and applications for cancer genomics.

Before joining Columbia, I was a Computer Science PhD student at UT Austin working with James Scott. My main research interests are health and wellness applications of machine learning, particularly those involving graphical models, Bayesian statistical methods, and scalable inference algorithms. I've worked on projects ranging from obesity and nutrition modeling to wearable fitness devices and large-scale multiple hypothesis testing for fMRI studies.

In a previous life, I was a software engineering researcher working with Eli Tilevich at Virginia Tech, where I got my BS and MS in Computer Science. My Master's thesis focused on inference techniques that learn transformation rules to automatically upgrade legacy applications to use the latest version of a given API. I've also co-founded a couple of startups and was a quant at a hedge fund.






































































Publications and Preprints

2018

Leaf-Smoothed Hierarchical Softmax for Ordinal Prediction
W. Tansey, K. Pichotta, and J. G. Scott
The 2018 AAAI Conference on Artifical Intelligence (AAAI-18), February 2018. [pdf] [code]

Maximum-Variance Total Variation Denoising for Interpretable Spatial Smoothing
W. Tansey, J. Thomason, and J. G. Scott
The 2018 AAAI Conference on Artifical Intelligence (AAAI-18), February 2018. [pdf] [code]

2017

Interpretable Low-Dimensional Regression via Data-Adaptive Smoothing
W. Tansey, J. Thomason, and J. G. Scott
The 2017 ICML Workshop on Human Interpretable, August 2017.

False Discovery Rate Smoothing.
W. Tansey, O. Koyejo, R. A. Poldrack, and J. G. Scott.
Journal of the American Statistical Association (JASA): Theory and Methods. [paper] [code]

Deep Nonparametric Estimation of Discrete Conditional Distributions via Smoothed Dyadic Partitioning
W. Tansey, K. Pichotta, and J. G. Scott
arXiv:1702.07398, January 2017. [preprint] [code]

2016

Diet2Vec: Multi-scale analysis of massive dietary data
W. Tansey, E. W. Lowe, and J. G. Scott
The 2016 NIPS Workshop on Machine Learning for Health, December 2016. [paper]

Multiscale spatial density smoothing: an application to large-scale radiological survey and anomaly detection.
W. Tansey, A. Athey, A. Reinhart, and J. G. Scott
Journal of the American Statistical Association (JASA): Applications and Case Studies, 2016. [preprint] [code] (currently undocumented but supported in the GFL package)

2015

A Fast and Flexible Algorithm for the Graph-Fused Lasso.
W. Tansey and J. G. Scott.
arXiv:1505.06475, May 2015. [preprint] [code]

Vector-Space Markov Random Fields via Exponential Families.
W. Tansey, O.-H. Madrid-Padilla, A. Suggala, and P. Ravikumar.
In International Conference on Machine Learning (ICML) 32, 2015. [pdf] [code]

(-\infty, 2014]

Accelerating Evolution via Egalitarian Social Learning.
W. Tansey, E. Feasley, and R. Miikkulainen.
The 14th Annual Genetic and Evolutionary Computation Conference (GECCO'12), Philadelphia, Pennsylvania, USA, July 2012. [pdf] [code]

Multiagent learning through neuroevolution.
R. Miikkulainen, E. Feasley, L. Johnson, I. Karpov, P. Rajagopalan, A. Rawal, and W. Tansey.
Advances in Computational Intelligence, pages 24-46, 2012.

Trailblazer: A Tool for Automated Annotation Refactoring.
M. Song, E. Tilevich, and W. Tansey.
An OOPSLA 2009 Tool Demo.

DeXteR - An Extensible Framework for Declarative Parameter Passing in Distributed Object Systems.
S. Gopal, W. Tansey, G. C. Kannan, and E. Tilevich.
In Proceedings of ACM/IFIP/USENIX 9th International Middleware Conference (Middleware 2008), 2008. [pdf]

Annotation Refactoring: Inferring Upgrade Transformations for Legacy Applications.
W. Tansey and E. Tilevich.
In The 2008 ACM SIGPLAN Conference on Object-Oriented Programming Systems, Languages, and Applications (OOPSLA 2008), October 2008. [pdf]

Efficient Automated Marshaling of C++ Data Structures for MPI Applications.
W. Tansey and E. Tilevich.
In Proceedings of the 22nd Annual IEEE International Parallel and Distributed Processing Symposium (IPDPS 2008), April 2008. [pdf]