Bill Sun

Interactive Resumé

About Me

Hi! I’m Bill Sun. I’m a first-year undergraduate student at Columbia University interested in pursuing a career at the crossroads of computer science, biomedical engineering, and healthcare. Currently, my main research interests involve the application of machine learning and deep learning techniques to improve human healthcare. I strive to integrate my knowledge of web development with my understanding of machine learning to create easily-accessible and accurate diagnostic processes to assist patients afflicted with various diseases. In my free time, I enjoy cooking and playing ping pong.
Bill

Education

Columbia University

Aug 2020 - May 2024
New York City, NY

B.S. Degree (In Progress)

GPA: 3.95
Enrolled in the Fu Foundation School of Engineering and Applied Science (SEAS). Intended Major: Computer Science
Activities and Societies: Columbia Robotics Club, Columbia Organization of Rising Entreprenuers (CORE), Columbia Economics Society

Georgia Institute of Technology

Aug 2018 - May 2020
Atlanta, GA

Dual Enrollment Program

GPA: 4.0
Selected Coursework: MATH 1554 (Linear Algebra), MATH 2551 (Multivariable Calculus), MATH 3012 (Applied Combinatorics), and MATH 2552 (Differential Equations)

Walton High School

Aug 2016 - May 2020
Marietta, GA

High School Diploma

GPA: 4.0
Activities and Societies: Math Honor Society, Science National Honor Society, Cryptocurrency Club, Chinese Language and Culture Society, Model UN, Walton Community Outreach

Skills

Languages and Frameworks

  • Python
  • HTML(5)
  • Javascript
  • React.js
  • MATLAB
  • Bootstrap
  • CSS
  • Node.js
  • Java
  • SQL
  • Angular.js
  • Vue.js

Experiences

Academic Research

Digital Video and Multimedia Lab

Aug 2020 – Jan 2021
Columbia University

Remote Research Intern

Worked on a project titled "Improving Few-Shot Object Detection through Attention-RPN and Enhanced Multi-Relation Detector." In the project, I implemented a Faster RCNN-based object detection algorithm on the 2014 COCO Dataset, and I developed an improved module for few-shot classification using PyTorch and Detectron2 in Python.

Bioinformatics and Bioimaging Lab

June 2019 – July 2019
Georgia Institute of Technology

Research Intern

Used Faster RCNN, You-Only-Look-Once, and other object detection algorithms to detect and classify various types of lesions (e.g. bone, liver, lymph) from assorted CT images using Keras and Tensorflow

Computer Science Lab

Aug 2018 - Aug 2020
University of Miami

Remote Research Intern

Compared performance of autoencoder-based neural network and feed-forward neural network in predicting stress and deformation of aortic valves under systolic pressure using MATLAB, Python, Keras, and Tensorflow

Tissue Mechanics Lab

June 2018 – July 2018
Georgia Institute of Technology

Research Intern

Used MATLAB to pre-process a collection of 176 aortic valve CT image sets, and utilized K-Means clustering algorithm in Python to obtain geometric means of valve annulus and sinus diameter of the patient dataset.

Tissue Mechanics Lab

June 2017 – Aug 2017
Georgia Institute of Technology

Research Intern

Used open-source machine learning codes and the direct linear transformation (DLT) algorithm to create a 3-D facial model from two 2-D images taken at different angles

Work Experience

Marietta Kumon Center

June 2019 - Jan 2020
Marietta, GA

Tutor/Grader

I tutored K-12 students in the Kumon math and reading program

Projects

turtle
StyleTransfer
StyleTransfer
bg img
Web application that utilizes a deep neural network to provide a real-time means of styling user images with various famous artwork styles (e.g. Van Gogh's Starry Night). The style transfer algorithm was taken from the tensorflow_hub library in Python. The frontend was built using HTML, CSS, and Javascript, and backend was built using Flask.
Live Demo!
malaria
X-Malaria
X-Malaria: Automated Malaria Diagnosis
bg img
Web application that performs malaria diagnosis from thin blood smear images using two deep learning models. A custom-trained You-Only-Look-Once (YOLO) object detection algorithm detects and segments individual red blood cells from the entire blood smear image, and a standard convolutional neural network classifies individual red blood cells as infected or uninfected by malaria parasite. The deep learning models were developed using Keras and Tensorflow in Python. The frontend was built using HTML, CSS, and Javascript, and backend was built using Flask.
Live Demo!
cat
HeartHelper
HeartHelper
bg img
Web application that diagnoses risk of heart disease from 13 clinical characteristics (e.g. sex, age, resting blood pressire) using the Gaussian Naive Bayes algorithm, which achieved ~90% diagnostic accuracy. The predictive models was developed using Sci-Kit Learn in Python. The frontend was built using HTML, CSS, and Javascript, and backend was built using Flask.
Live Demo!
turtle
Summarizr
Summarizr: Extractive Text Summarization
bg img
Developed a graph weighting-based extractive text summarizer using the NLTK and networkx packages in python; also implemented as a chrome extension. Summarizr won "Best ML Hack" at Columbia’s DevFest '21 Hackathon
Live Demo!
malaria
ConditionalGAN
Conditional GAN - Fashion MNIST
bg img
Tensorflow implementation of a unconditional generative adversarial network (GAN) and a conditional GAN trained from 80 epochs on the fashion MNIST dataset. The conditional GAN takes in an additional class label as input, and outputs an image in the corresponding class. All output images are uniquely generated.
Link to git repo
cat
ClimateTracker
ClimateTracker
bg img
Developed a web application that provides clean, understandable visualizations of historical climate data and energy consumption by each state for the last 50 or so years, as well as forecasts for future surface temperatures by state for the next 100 years. Scraped the top 10 most recent news articles regarding climate change and clean energy, and through natural language processing techniques, provides auto-generated summary and list of keywords
Live Demo!

Contact

Where to Find Me:
1107 Midsummer Court NE, Marietta, Georgia, USA
Email Me At:
bys2107@columbia.edu
Let's Connect:
My Linkedin Profile