Research
My
postdoctoral research is at the intersection of machine learning and
ecology. The unifying theme of my current work is incorporating
nonparametric methods like boosted regression trees into ecological
latent variable models. Since observations of ecological
processes are often imperfect and incomplete, ecologists often use
hierarchical latent variable models to specify a model of the true
ecological process of interest along with a model relating the
observations to that process. However, ecological processes are
also complex, and the need to identify a parametric function of
potentially many input variables for use in a hierarchical latent
variable model may be overwhelming. Regression trees avoid the
need to define the relationship of the inputs to the model quantities
in advance, and they automatically capture nonlinearities and
interactions. Unfortunately, they do not handle latent variables,
so they cannot take advantage of the expressiveness of hierarchical
models. My goal is to unite these two approaches, in hopes that
ecologists will not have to choose between an ecologically plausible
hierarchical model and the ability to deal with many inputs with
complex effects.
I am a SEES
(Science, Engineering, and Education for Sustainability) Fellow,
funded by the National Science Foundation. I am thrilled to be
working with two excellent mentors: Dr.
Thomas G.
Dietterich
and Dr.
Matthew G. Betts.
Software
We plan to
release an R package implementing the methods we are
developing. That package is under development, but check back
here for updates!
Education
Ph.D.,
Computer Science Department, 2009
Carnegie Mellon
University, Pittsburgh, PA
Advised by Dr.
Tom
M. Mitchell
B.S.,
Computer Science & Engineering, 2002
Bucknell
University, Lewisburg, PA
Refereed Publications
"Species
distribution modeling for the people:
Unclassified landsat TM imagery predicts bird distributions at fine
resolutions in forested landscapes,"
S. Shirley, Z. Yang, R.A.
Hutchinson, J. Alexander, K. McGarigal, and M.G. Betts,
Diversity
and Distributions, in
press.
"Project
and Analysis Design for Broad-Scale Citizen
Science,"
W. Hochachka, D. Fink, R.A. Hutchinson,
D. Sheldon, W-K.
Wong, and S. Kelling,
Trends in Ecology
and Evolution, Vol. 27, Issue
2, pp. 130-137.
February 2012.
"Incorporating Boosted
Regression Trees into Ecological
Latent Variable Models,"
R.A. Hutchinson, L-P. Liu, and T.G.
Dietterich,
Proceedings of the
Twenty-fifth Conference
on Artificial Intelligence, 2011.
"Modeling Experts
and Novices in Citizen Science data for
Species Distribution Modeling,"
J. Yu, W-K. Wong, and R.A. Hutchinson,
International Conference on Data Mining,
2010.
"Modeling
fMRI data generated by
overlapping cognitive
processes with unknown onsets using Hidden Process Models,"
R.A.
Hutchinson, R.S. Niculescu, T.A. Keller, I. Rustandi, and T.M.
Mitchell,
Neuroimage (2009),
doi:10.1016/j.neuroimage.2009.01.025. (preprint version)
"Hidden Process Models,"
R.A.
Hutchinson, T.M. Mitchell, and I. Rustandi,
International
Conference on Machine Learning 2006. June 2006.
"Learning to Decode
Cognitive States from Brain Images,"
T.M. Mitchell, R.A.
Hutchinson,
R.S. Niculescu, F.Pereira, X. Wang, M. Just, and S. Newman,
Machine
Learning, Vol. 57, Issue 1-2, pp. 145-175. October 2004.
"Training fMRI Classifiers
to Detect Cognitive States across Multiple Human Subjects ,"
X.
Wang, R.A. Hutchinson, and
T.M. Mitchell,
Neural Information
Processing Systems 2003. December 2003.
"Classifying Instantaneous
Cognitive States from fMRI Data,"
T.M. Mitchell, R.A. Hutchinson,
M.
Just, R.S. Niculescu, F. Pereira, and X. Wang,
American Medical
Informatics Association Symposium, October 2003. (received Best
Foundational Paper Award)
"Reducing Boundary Friction
Using Translation-Fragment Overlap,"
R. Brown, R.A. Hutchinson, P.
Bennett, J.G. Carbonell, and P. Jansen.
Machine Translation Summit IX. 2003.
Other Papers
"Machine
Learning for Computational
Sustainability,"
T.G. Dietterich, E. Dereszynski, R.A.
Hutchinson,
and D.
Sheldon,
International
Conference on Green Computing (IGCC),
2012.
"Modeling
Experts and
Novices in Citizen Science data for
Species Distribution Modeling,"
J. Yu, W-K. Wong, and R.A. Hutchinson,
OSU EECS Technical Reports, October
7, 2010.
"Hidden Process
Models,"
R.A. Hutchinson,
Ph.D.
Thesis, CMU
Technical Report
CMU-CS-09-179. December 18, 2009.
"Hidden Process
Models,"
R.A. Hutchinson,
Thesis
Proposal, May 26, 2006.
"Hidden Process Models,"
T.M.
Mitchell, R.A. Hutchinson, and
I. Rustandi.
CMU Technical Report
CS-CALD-05-116. February
17, 2006.
"Maximal Lattice
Overlap in Example-Based Machine Translation,”
R.A. Hutchinson,
P.N.
Bennett, J. Carbonell, P. Jansen, and R. Brown.
Carnegie Mellon
University Technical Report
CMU-CS-03-138.
June 6,
2003.
Invited Talks
"Machine
Learning for Ecological Science and Environmental Policy", with
T.G.
Dietterich and D. Sheldon,
tutorial at the International
Conference on Machine
Learning (ICML), June 28, 2011.
(slides)
"Machine Learning Problems in Species Occupancy Modeling,"
SCHARP Brown Bag Series, Fred Hutchinson Cancer Research Center,
Seattle, WA. March 25, 2010. (slides)
"Hidden Process Models with Applications to fMRI Data,"
Biostatistics and Bioinformatics Seminar, Fred Hutchinson Cancer
Research Center, Seattle, WA. March 24, 2010. (slides)
"Hidden Process Models with Applications to fMRI Data,"
Topic Contributed Session, Joint
Statistical Meetings (JSM), Washington, DC. August 2,
2009. (slides)
Other Talks and Posters
"Site occupancy
models with regression trees (OD-BRT): A comparison
with standard site occupancy models (OD) and boosted regression trees
(BRT)",
Ninety-seventh Annual
Meeting of the Ecological Society of America (ESA), Portland,
OR, August 9, 2012.
"Incorporating
Boosted Regression Trees into Site
Occupancy
Models,"
AVES Seminar,
Department of Fish and Wildlife, Oregon State University, Corvallis,
OR. November 17, 2011. (slides)
"Incorporating Boosted
Regression Trees into Ecological
Latent Variable Models,"
Twenty-fifth
Conference
on Artificial Intelligence (AAAI), San Francisco, CA.
August 11, 2011. (slides)
"Combining Boosted Regression Trees and Hierarchical Species Occupancy
Models," (talk and poster),
International Conference on Computational Sustainability, Massachusetts
Institute of Technology, Cambridge, MA. June 27, 2010. (poster)
"Parameter Estimation in a Hierarchical Model for Species Occupancy," (poster
with T.G. Dietterich),
Neural
Information Processing
Systems (NIPS) Workshops: The Generative and Discriminative
Learning Interface, Whistler, BC. December 12, 2009.
(abstract)
"Modeling
fMRI data generated by overlapping cognitive processes with unknown
onsets using Hidden Process Models," (poster with T.M. Mitchell),
Statistical Analyses of Neuronal Data
Workshop (SAND) 2008, Pittsburgh, PA.
May 30, 2008. (slides)
"Hidden Process Models for Analyzing fMRI Data,"
Graduate Student Seminar Series, Carnegie Mellon School
of Computer
Science. May 11, 2007. (slides)
"Hidden Process
Models: Decoding Overlapping
Cognitive States with Unknown Timing,"
Neural Information Processing
Systems (NIPS) Workshops: New
Directions on Decoding Mental States from
fMRI Data, Whistler, BC. December 8, 2006. (slides)
"Hidden Process Models,"
Women in Machine Learning
Workshop (WIML), San Diego, CA. October 4, 2006. (slides)
"Hidden
Process Models,"
International Conference on Machine Learning (ICML) 2006, Pittsburgh, PA. June
28,
2006. (slides)
"Learning
to Identify Overlapping and Hidden
Cognitive Processes from fMRI Data,” (poster with T.M. Mitchell
and I. Rustandi),
Human Brain
Mapping (HBM) 2005, Toronto, ON. June
2005. (slides)
“Using Hidden Process Models to Decode
Cognitive
States from fMRI Data,”
Brain
Science Seminar, Carnegie Mellon
University. April 2005.
"Hidden Process Models for Body Monitoring Data,"
BodyMedia, Inc., Pittsburgh, PA. March 2005. (slides)
Teaching
Co-advisor (with
Tom Mitchell) for an undergraduate senior honors
thesis, 2006-2007 academic year
TA for Artificial
Intelligence (undergraduate level), Spring 2007
TA for Machine
Learning (undergraduate/master's level), Fall
2006
Service
Board Member (Committee Coordinator) for the Oregon State Postdoctoral
Association (2012-2013).
Program Committee Member for
the
1st annual
NorthEast Student Colloquium on Artificial Intelligence (NESCAI) 2006.
Journal reviewing: NeuroImage, Computational Neuroscience,
Ecology, Ecosphere.
Conference/workshop reviewing: International Conference on Machine
Learning (ICML), NorthEast Student Colloquium on Artificial
Intelligence (NESCAI), Women in Machine Learning Workshop (WiML).
Grant reviewing: National Science
Foundation (NSF).
Personal
I enjoy ultimate
frisbee, racquetball, knitting, and cooking. My
husband and I
have two dogs, Luna and Sola, who never
fail to turn our
hiking/camping trips into adventures.
