Jordan Hoffmann

Education:

PhD Student in Applied Mathematics at Harvard University (Expected: 2019)
MACHINE LEARNING IN PHYSICS
◦ Studying the degree of order in a hallmark disordered system.
◦ Work closely with experimentalists to create testable predictions.
MORPHOGENESIS PROJECT
◦ Use computational tools to quantitatively characterize early insect development.
◦ Begin by fusing hundreds of different images to create a 3-D reconstruction of a developing cricket embryo through time.
◦ Using ML, segment the 3-D images into nuclei and background, then track the motion of nuclei through time and space allowing for divisions.
◦ Characterize the motion of nuclei and make predictions which can be experimentally verified.
◦ Develop physical models describing the motion of nuclei in the embryo.
GEOMETRIC STUDY OF INSECT WINGS
◦ Developed a highly customizable segmentation tool to segment a large variety of wing images.
◦ Analyzed the geometric features of nearly 1000 insect wings.
MS Applied Mathematics, Harvard University 2015
BS Physics and Mathematics, Johns Hopkins University 2014

Experience:

PhD Student in the Rycroft Lab, Harvard
Joint Genome Institute, Summer 2016, Supervisor: Zhong Wang
Lawrence Berkeley National Lab, Summers 2015 and 2017
Undergraduate Researcher, Johns Hopkins University 2010-2014
Undergraduate Researcher, Rush Medical School Summer 2011 and 2012

Publications:

See Google Scholar

Awards:

Department of Energy Computational Science Graduate Fellowship (DOE CSGF)
James Mills Peirce Scholarship, Harvard Uiversity
QuantBio Student Award, Harvard Uiversity
Simmons Award, Harvard Uiversity
"BEST USE OF THE BLOOMBERG API" and "BEST FINANCIAL DATA HACK" at HackMIT with Matt Sheckells
Woodrow Wilson Undergraduate Research Award, Johns Hopkins University

Coding:

Happily use Python, C++, Mathematica
Have used Scala/Spark, Fortran, AWS
Happily use TensorFlow, used Theano quite a bit, but a couple years ago. Used Spark for a summer.

Other:

I teach chess at an elementary school for two hours, once a week.
I helped with this CarTalk article (NPR). As these things go, I played no part in the writing or interpretation of the data.