Seungjae Ryan Lee

Math Major, CS & ML Minor

Resume

About Me


I am a math major at Princeton University with minors in computer science and machine learning. I believe in learning by doing, and I always welcome new, exciting challenges that will help me grow. Each summer, I explored different interests, from conducting math research, founding a startup, and interning as a software engineer or a machine learning engineer. I also developed multiple personal projects and contributed to several open-source projects, I spend my free time teaching or translating, because I believe that sharing knowledge will make the world a better place.

Education


Princeton Logo

Princeton University

Bachelor of Arts in Mathematics - -

  • Minors in Computer Science and Machine Learning
  • Relevant Coursework: Algorithms, Probability, Statistics, Computer Vision, Seminar on Math for Data Science, Graph Theory, Intro to Computer Systems, Advanced Programming Techniques, Combinatorics, Topology, Complex Analysis, Abstract Algebra

Publications


Teaching Icon

Improving Current and Future Offerings of a Data Science Course through Large-Scale Observation of Students

11 authors including Seungjae Ryan Lee

Accepted to SIGCSE 2021

  • Discovered that enthusiasm had a substantial impact on students' performance
  • Formulated actionable conclusions for instructors to improve introductory data science courses
MagNet Icon

MagNet: A Machine Learning Framework for Magnetic Core Loss Modeling

Haoran Li, Seungjae Ryan Lee, Min Luo, Charles Sullivan, Yuxin Chen, and Minjie Chen

Accepted to IEEE COMPEL 2020

  • Create a dataset of voltage and current data of magnetic components under different waveforms measured in the real world
  • Train machine learning models to estimate the magnetic core loss
View Publication
Markoff Graph

Experiments with the Markoff Surface

Matthew de Courcy-Ireland and Seungjae Ryan Lee

Published in Experimental Mathematics (2020)

  • Analyzed graph structures and spectral patterns of Markoff graphs with four million vertices and six million edges
  • Proved a deterministic formula for the number of orbits of any Markoff graph in a prime field
View Publication

Experience


Lighting Icon

Princeton Power Electronics Research Lab

Research Assistant

-

I worked with Haoran Li under Minjie Chen to create a dataset of a large amount of voltage and current data of different magnetic components measured in the real world. With the collected data, we experimented with combinations of existing techniques and deep learning algorithms to predict magnetic core loss.

  • Develop neural networks with PyTorch to predict magnetic core loss under arbitrary excitations
  • Perform hyperparameter optimization with Optuna to find optimal neural network architecture
Python PyTorch PyTorch Lightning Optuna Weights & Biases
Teaching Icon

Princeton University

Course Assistant - SML201: Introduction to Data Science

-

I have worked as a course assistant for two semesters for an introductory data science course, leading class discussion in weekly meetings (precepts).

  • Teach data science and R to undergraduate students enrolled in SML201: Introduction to Data Science
  • Study and evaluate pedagogical approaches to teaching data science in online and in-person settings
R
Bloomberg Logo

Bloomberg L.P.

Software Engineer Intern

-

I interned in the Data Technologies Automation Engine team and used unsupervised learning and weakly supervised learning to improve performance without using labelled data.

  • Trained a word2vec word embedding model using Bloomberg's financial data
  • Reduced the size of the embedding by 97% and increased inference speed by 5 times while maintaining performance
  • Used weakly supervised learning to train a generative model for multi-entity relation extraction
Python Gensim Snorkel
SKT Logo

SK T-Brain

Machine Learning Research Intern

-

I have interned in the Conversational AI team and participated in the Eighth Dialog System Technology Challenge (DSTC8). My work focused on developing a word-level policy that maps a context to a response.

PyTorch Python Microsoft Azure JavaScript VueJS
TensorFlow Logo

Google Summer of Code: TensorFlow

Student Software Developer

-
-

My Google Summer of Code proposal was accepted by the TensorFlow team in Summer 2019 and 2020. In 2019, I worked on TF-Agents, TensorFlow's reinforcement learning library. In 2020, I worked on Swift for TensorFlow.

  • Implemented and documented popular reinforcement learning algorithms (DQN, PPO) using Swift for TensorFlow
  • Prototyped Random Network Distillation, a bonus-based exploration reinforcement learning algorithm
TensorFlow TensorBoard Python Google Cloud Platform Swift for TensorFlow
Scratchwork Logo

Scratchwork LLC

Co-founder and Software Developer

-

I founded Scratchwork with 3 co-founders offering a whiteboard app customized for communication between researchers.

  • Designed and built the main dashboard page allowing users to create, edit, or delete whiteboards
  • Implemented login with Google OAuth integration using Passport.js
JavaScript MongoDB ExpressJS NodeJS ESLint

Projects and Contributions


Optuna Logo

optuna/optuna

A hyperparameter optimization framework

python machine-learning hyperparameter-optimization
KAIR Logo

kairproject/kair_algorithms

Reinforcement learning for robot control

python reinforcement-learning robotics
My Image

seungjaeryanlee/MagNet

Database for predicting magnetic core loss with neural networks trained with AutoML

machine-learning modeling magnetics dataset
My Image

seungjaeryanlee/retro-agents

Reinforcement learning agents trained on Sonic the Hedgehog

python reinforcement-learning

Volunteer


Reproducibility Challenge Logo

Reproducibility Challenge

Reviewer

COVID Translate Project Logo

COVID Translate Project

Korean-to-English Translator

I translated parts of Korean CDC's COVID-19 Response Guidelines under Professor Sebastian Seung.

Princeton Logo

Princeton University

Undergraduate Course Assistant

I hold weekly precepts and one-to-one appointments for SML201: Introduction to Data Science.

NYU Logo

New York University

English-to-Korean Translator

I translate parts of the course materials for Deep Learning (DS-GA 1008) to Korean.

OpenMined Logo

OpenMined

English-to-Korean Translator

I translated parts of the OpenMined's PySyft tutorial to Korean.

Modulabs Logo

Modulabs

Deep Reinforcement Learning Seminar Leader

I developed the syllabus and led the Deep Reinforcement Learning Seminar with 11 participants.

Modulabs Logo

Deep Learning Zero To All 2

PyTorch Content Contributor

I wrote Jupyter notebooks and recorded lectures for Korean introductory deep learning course in PyTorch.

Contact