This project summarized latent topics from COVID-19 related scientific articles and created a hierarchical framework for topic-based search by modifying traditional Non-Negative Matrix Factorization (NMF) method. We also implemented a website for scientists' investigation. This paper was accepted by EMNLP 2020 (Paper upon Clicked the "Full Story" button).
Using CSS, HTML & JavaScript
This project is my favorite one so far for the Stack, UCLA Daily Bruin. During the COVID outbreak in April, several students and I examined how quickly COVID-19 could spread throughout the undergraduate student body if UCLA would reopen at 2020 Fall Quarter via designing UCLA student network model.
This project shows my 2019 summer internship for Gang Reduction and Youth Development (GRYD) program at the City of Los Angeles. We used machine learning algorithms and mathematical modeling to evaluate and imrpove its effectiveness.
This project compares the performances of machine learning algorithms when dealing with categorical variables. It was invited by 2019 Symposium on Data Science & Statistics for poster presentation, and is preparing for the manuscript publication.
"How to Make your Video Trend on Youtube?" Four students and I from UCLA DataRes Club explored how important factors, such as growth in the number of views and time posted, make videos trending by analyzing the Youtube data on Kaggle.
Using CSS, HTML & JavaScript
"Walking to Class" is my first project for the Stack, UCLA Daily Bruin, which is the biggest social media club ran by students. This project helps students explore new routes from the dorms to campus and provide various statistics about each route
This project is my own version of the very popular game 2048. It calculates the total score and also determines whether you win or lose. It is developed in C++ and uses Qt to make beautiful UIs. You can download and play it on your own computer.
This project guided the preparation of registered reports for successful publication by researched over 100 journals’ policies of registered reports. I led a group of 4 and we together developed predictive analytics policies for journals.