Chris Cha is trying new things
i’m a first year master’s student at UC San Diego, currently focusing on the intersection between machine learning and human-computer interaction. i’m mainly interested in data science, natural language processing, and human-centered healthcare.
i like retro hardware and i think that moray eels are cool. i also consume a lot of media so if you found something neat to watch or read, you should message me!
this website is a work in progress. i’m still learning React and webdev is hard!
Neat Projects
Spoiler Detection Model - 02/2022
This was a component of a hackathon that would classify text as a spoiler and flag the text. It was a binary classification model using DistilBERT, and it achieved an impressive 83% accuracy on our test set without any other context. This was then hosted using Flask to host the model.
Tech used: Pytorch, Scikit Learn, Pandas, Scipy, MongoDB, JS, Flask
Anilist Rating Prediction - 12/2021
A course project for my recommender systems class. My main role was creating a dataset collection pipeline to scrape and process entries in preparation for model training, eventually ending up with a feature vector of 13 features within our dataset to describe the content.
Tech used: Python, GraphQL, Pandas, Scikit Learn, Scipy
Music Genre Transfer with CycleGAN - 07/2021
Implemented a CycleGAN‑based model to transfer musical style from jazz to classical using Tensorflow and Keras. Preprocessed data using Numpy by sampling MIDI files to discrete time and creating a matrix representation of time steps and pitches.
Tech used: Python, Tensorflow, Keras
Ventilator Webjack Tester - 12/2020
This was an early static site we used to test and debug a ventilator’s modem systems. It works kind of like dial-up internet, where a series of audio waveforms transfer information. Now I use it to make neat sound effects. Be careful if you use it because it can be quite loud.
Tech used: Javascript, JQuery, HTML, CSS, Pages
Project #X - XX/XXXX
Lorem ipsum dolor sit amet, consectetur adipiscing elit. Ut posuere in tortor in cursus. Aliquam iaculis enim sit amet diam pharetra volutpat. Etiam male.
Tech used: Proin Rhoncus enim sit amet
Publications
Tom Corringham, Daniel Spokoyny, Eric Xiao, Christopher Cha, Colin Lemarchand, Mandeep Syal, Ethan Olson, Alexander Gershunov (2021). “BERT Classification of Paris Agreement Climate Action Plans.” Tackling Climate Change with Machine Learning Workshop at ICML 2021.
Xiaotong (Tone) Xu, Khuyen Le, Christopher Cha, Samuel Fleet, Steven P. Dow (2021). “Crowdsourcing Exploratory Cues for Idea Browsing and Inspiration Discovery.” HCOMP 2021.