(mailing address, phone, or a pdf copy of this resume available on request)
July 2015-December 2015
Stanford Virtual Human Interaction Lab
September 2012-June 2015
Aided the development of research projects by Ph.D students in the lab under the supervision of Professor Jeremy Bailenson. Specifically, aided in the development of virtual worlds that would be the settings of experiments and demos, created a custom playback system for Kinect recordings of human movements, assisted with various machine learning tasks, and helped develop a Unity application which would allow Android phones to be used as head-mounted devices for virtual reality.
Data Science Intern
The Backplane, Inc.
Created the predictive analytics system incorporating a Mahout and MapReduce system, which predicted the user retention rate of a website with 20,000 active users, and created a custom AB split testing solution for their AB testing needs. Presented the architecture of this system to coworkers.
Stanford Symbolic Systems Program
September 2013-June 2015
Advised students about research, classes, and life. Inducted over two dozen new students into the Symbolic Systems program. Planned field trip to places like Palantir and Facebook. Helped with preparations for the departmental Distinguished Fellow award for 2014 to Douglas Hofstadter. Led student preparations for Distinguished Fellow award for 2015 to Edward Snowden, including making arrangements to contact the awardee.
2014 - 2015
Master of Science
2011 - 2015
Bachelor of Science
spec. in Artificial Intelligence
Extremal Optimization for Boltzmann Machines Usually, fully connected Boltzmann machines are not suitable for use in real applications because the simulated annealing takes too long. I tested what happens if you don’t use simulated annealing, and found about one order of magnitude speedup over simulated annealing.
For Psychology 209, Neural Nets
Bak-Tang-Wiesenfeld Text Generation Used a physics-based cellular automaton to generate Markov chain-style text. Nobody had done this before.
Not for school.
2nd place at the 2013 Stanford Winter ACM Hackathon.
Alpha-Phi Algorithm Novel algorithm and implementation in MapReduce developed in a team for the class, Project in Mining Massive Data Sets, under Jeffrey Ullman. 3% better AUC than the published Cascades algorithm for predicting the size of a Twitter cascade.
I have been paid to use
I also know
C, C++, Scala, Erlang, Mathematica, Scheme
HTML, CSS, LaTeX Linux, Windows and AWS administration