Hello!
I am Ying-Chi.
Computer Science @ University of Michigan
About Me
Get to know me!I have 3+ years of experience in software engineering and development.
My name is Ying-Chi (Brian) Chen and I'm a senior Computer Science student at the University of Michigan. I am curious, passionate and determined, and I love learning new skills. I excel in collaborative team environments, but I also enjoy a bit of competition. I am proficient in multiple programming languages and experienced with writing scalable, high-quality code and unit tests with good test coverage.
Languages:
- Python
- C++
- JavaScript
- Java
- SQL
- R
- Git
Libraries:
- OpenAI api
- TensorFlow
- PyTorch
- OpenCV
- Pandas
- NumPy
Tools/Technologies:
- React
- Flask
- PostgreSQL
- MongoDB
- AWS EC2
- AWS Lambda
- Microsoft Azure
Education
My academic history!I pride myself in learning from experts and professionals.
Experience
My professional history!I love designing innovative solutions and solving complex problems.
Portfolio
Check out my work!NLP Reddit Forum Classifier
Natural language processing model that classifies reddit text speech into happy, sad, and neutral moods.
View ProjectMicrosoft Azure Web Calculator
Python calculator web application built with Django, containerized with Docker and hosted on Azure App Service with requests data stored in Azure PostgreSQL flexible database.
View ProjectGAN Anime Face Generator
Created a deep learning face generator that generates fictional anime character faces using Pytorch. The generator model was built using a transposed convolutional neural network and the discriminator model was built using a deep convolutional neural network.
View ProjectSimulation of Pipelined RISC-V CPU
Implemented a complete simulation of a 8-bit pipelined CPU that runs on the LC-2K instruction architecture. The simulator program combines the assembler, linker, pipeline, and cache to simulate the LC-2K CPU.
View ProjectResearch
My research experience and publications!Utilized Python libraries like OpenCV, Scikit-Image, and NumPy to preprocess 1000+ fundus images from different hospitals, including methods like resizing, noise-reduction, contrast enhancement, and feature-extraction.
Fine-tuned an EfficientNet-B3 convolutional neural network model to achieve 95% accuracy on test dataset and improved classification accuracy on validation data from 80% to 92.5%.