No longer on the market
I joined Expedia Group in Spring 2018 to work on machine learning and automated bidding.
Feel free to peruse my publications and experience.
Publications
Education
Thesis:
Kernels and Geometry of Machine Learning
Emphasis:
Machine Learning, Algorithms, Theory
2017
Salt Lake City, UT, USA
1999
Salt Lake City, UT, USA
Awards
Poster, Kernels and Geometry of Machine Learning.
2016
SDM Doctoral Forum
Miami, FL, USA
Experience
Applied Researcher, Global Marketing Organization
Developing automated bidding algorithms for optimal ad placement
2018-current
- Developed machine learning algorithms and software for SEM
Senior Software Architect
Maintained fulfillment portal software for wine and liquor distributors
2018
Staff R&D Engineer
Extended and maintained open source toolkit SMQTK (for machine learning, image processing, and data processing)
2016-2018
- Research: Researched graph learning algorithms for “patterns of life” analysis
- Python: Added new features (nearest neighbor data structures such as Facebook's FAISS) and web service frontends
Research Assistant
Advisor: Suresh Venkatasubramanian
2009-2016
-
Deep learning: Multilayer perceptrons, convolutional neural networks, working with kernel methods in deep learning applications
-
Supervised machine learning: support vector machines (SVM), multiple kernel learning (MKL), kernel methods, optimization
-
Computational geometry: non-Euclidean geometry, Bregman geometry, geometric approximation algorithms
-
C++, Python: Developed and maintained software implementing our MWUMKL algorithm. C++ implementation, Python front-ends for general use and integration with scikit-learn ML package, Swig interface layer.
Teaching Assistant, Clustering
Instructor: Suresh Venkatasubramanian
Fall 2016
Teaching Assistant, Advanced Algorithms
Instructor: Suresh Venkatasubramanian
Fall 2014
Google
Software Engineering Intern
Host: Shankar Kumar
Research: Knowledge Spam (Burst Detection)
Summer 2013
- Researched methods of timeline burst detection for knowledge topics (events, personalities, products, etc.)
- C++: Developed convolution implementation of continuous wavelet analysis using Mexican hat wavelets
- C++, Flume: Used Flume distributed computation framework to process massive amounts of topic information
- Data-mined appropriate settings for finding bursty topics
Software Engineering Intern
Search Quality, Localization
Summer 2010
- C++, MapReduce: Wrote custom filters and data-mined web for location data
Onyx Graphics
Software Engineer
Implementing new features, maintenance
2004-2009
- JNI (C++, Java): Developed and maintained interface between
legacy large-format printing code for backend and Java application front-end using JNI
- Native acceleration: Added SSE instructions to accelerate
processing of large-format printing kernel
- Halftoning technologies
- Investigated new technologies
Cabletron Systems/Enterasys Networks
Firmware Engineer
Ported and maintained Unix-based routing software on real-time embedded platform
1999-2002,2003
- C, C++: Backported ordinary routing protocols (OSPF, BGP, RIP) and multicast routing protocols (PIM-SM, DVMRP) to proprietary platform
Moog Aircraft
Tech Writer, Test Engineer
Updated specifications (MS Word) and tested software (MS Excel, various)
2002-2003
Posts