I'm a MS/BS student at Stanford with a concentration in machine learning and a minor in creative writing. I'm currently working in Chris Ré's group, using weak supervision to make deep learning more accessible in domains like medical imaging and computer vision.
Most recently, I worked as an engineer at Sift Science fighting fraud with machine learning. In the past, I've also taken on product roles, designing VR experiences at Xbox and building point-of-care healthcare tools at EMGuidance. In 2016, I had a lot of fun co-directing TreeHacks, Stanford's largest intercollegiate hackathon.
I like writing to explain things clearly. One of these efforts is Paths, a series I created on YC's blog to make technical fields more accesible.
Vincent Chen, Paroma Varma, Madalina Fiterau, James Priest and Christopher Ré. Machine Learning for Health (ML4H) Workshop, Neural Information Processing Systems (NIPS), 2017.
Using weak-supervision, we learn probabilistic training labels for aortic valve MRIs.
Vincent Chen, Paroma Varma, Madalina Fiterau, James Priest and Christopher Ré. Medical Imaging (MED-NIPS) Workshop, Neural Information Processing Systems (NIPS), 2017.[pdf]
Vincent Chen*, Dan Yu*. CS229.
We developed a new method to predict demographics based on FourSquare data by engineering features based on check-ins mapped to U.S. census tracts.
Vincent Chen*, Liezl Puzon*, Christina Wadsworth*. CS231N.
For the superresolution task, we proposed several methods to introduce auxiliary, conditional information into generative adversarial networks (GANs) that produce images more tuned to the human eye.
Vincent Chen*, Liezl Puzon*, Eduardo Torres Montaño*. Advisor: Danqi Chen. CS224N.
We implemented several approaches for sequence-to-sequence summarization on the CNN/DailyMail dataset using attention mechanisms and pointer networks.
Vincent Chen*, John Kamalu*, Scott Kazmierowicz*.CS221.
We introduced an adversarial twist to the classic game, Snake, by allowing the food to move (hence, mice) and implemented an agent to play the new game.
Vincent Chen, Geza Kovacs. Advisor: Michael Bernstein. Research, Stanford HCI.
We created visualizations of academic curricula, where nodes in each curriculum network graph were modules for on-demand learning.