Polypilot product mascot

Introducing PolyPilot:

Our AI-Powered Mentorship Program

Start your trial today

Learn More
profile picture

Kyle G

- Research Program Mentor

PhD candidate at University of Michigan - Ann Arbor

Expertise

Nonconvex optimization, machine learning, signal processing, online and scalable algorithms

Bio

I'm a Ph.D. candidate of the Signal Processing Algorithm Design and Analysis (SPADA) lab advised by Prof. Laura Balzano at the University of Michigan. My research focuses on nonconvex optimization problems with applications in computer vision, signal processing, medical imaging, environmental sensing, data science, and more. My favorite problems include low-rank matrix and tensor factorization, missing data completion, scalable online algorithms, heteroscedastic models, optimization on Riemannian manifolds, and randomized algorithms. When I'm not doing research, I love practicing piano or the guitar, playing beach volleyball in our LGBTQ league, floating on the river, or watching Netflix.

Project ideas

Project ideas are meant to help inspire student thinking about their own project. Students are in the driver seat of their research and are free to use any or none of the ideas shared by their mentors.

Movie Titles Recommendation Engine

Modern day streaming platforms (e.g. Netflix, Hulu, Disney+, etc.) seek to model user preferences for movies/TV shows in order to better recommend new titles. If we collect a matrix of users to movie titles, with the entries being the user ratings of the viewed titles (e.g. 1-5 "stars"), the matrix is highly incomplete, i.e. most of the entries (90-95%) are missing since most viewers have only watched a handful of titles on the platform. With so much missing data, it might seem hopeless. Nevertheless, we can build algorithms to learn a low-dimensional model that allows us to complete and predict the missing entries. From this model, we can also learn factors that explain user preferences for titles (e.g. does user X prefer rom-coms? Marvel movies?). How accurately can we predict user ratings based on the algorithm we design? Can we cluster users together based on their preferences? Can we build "online algorithms" that update their estimates from the addition of new users, rather than computing the entire model from scratch each time?

Coding skills

Python, MATLAB, Julia

Teaching experience

I previously served as a graduate student instructor for the Fall 2020 offering of EECS 505 (Computational Data Science and Machine Learning) at the University of Michigan, and was recognized with an ECE GSI Honorable Mention. In addition, I occasionally meet with undergraduate/graduate students to advise on research projects.

Credentials

Work experience

Sandia National Laboratories (2021 - 2021)
Scalable Modeling & Algorithms Graduate Research Intern

Education

University of Wyoming
BS Bachelor of Science (2017)
Electrical Engineering
University of Michigan - Ann Arbor
MS Master of Science (2019)
Electrical and Computer Engineering - Computer Vision
University of Michigan - Ann Arbor
PhD Doctor of Philosophy candidate
Electrical and Computer Engineering - Systems

Interested in working with expert mentors like Kyle?

Apply now