Projects
Covey.Town Badges
What is Covey.Town: Covey.Town is an open-source virtual meeting space designed as a 2D game. Users can log-in to explore various interactable areas and collaborate with others through video calls.
I collaborated with teammates to develop a new feature for Covey.Town. The new feature allows users to earn achievements, which are represented through displayable badges. These badges are earned in real-time, which means that users will receive their badge as soon as an achievement's criteria is met; this also means that all users are able to see other users' badges. When a user exits Covey.Town and logs in again, they will keep their hard-earned badges. This is possible because we implemented a PostgreSQL database that keeps track of users' in-game data, which includes badges that have already been earned.
Made with: TypeScript, ReactJS, PostgreSQL, Express, Jest
FreeCell Solitaire Game
What is FreeCell: FreeCell is a solitaire card game played using the standard 52-card deck with all cards dealt face-up.
I programmed FreeCell as an interactive text-based game and implemented Object-Oriented programming principles to create a Model, Controller, and View. Users type their desired move in the format: origin pile, card index to move, the destination pile. The game checks if the move is valid; if the move is not allowed, a specific error message will inform the user and suggest to try again. In order to achieve proper program behavior, I performed rigorous testing to ensure that all functions work properly. The game went through various iterations where new features were added on top of the pre-existing code base.
Made with: Java, JUnit
NBA Most Improved Player (MIP) Award Prediction
How the award is decided: Every season, the National Basketball Association's (NBA) Most Improved Player (MIP) award is given to one player. A panel of sportswriters each have a first-place, second-place, and third-place vote, worth 5, 3, and 1 point, respectively. When votes are cast, nominees accrue voting shares, representing the percentage of votes that they received. The player with the greatest number of voting shares wins the award.
I worked with datasets and machine learning models (K-nearest neighbors, Ridge, Lasso) to try and predict the voting shares that each player will receive in a season. I extracted datasets of thousands of players' statistics to clean and feed into the machine learning models; a portion of the data would be used for training the model, and another portion would be used for testing. Further tuning was applied to the model to produce better results: GridSearchCV for hyperparameters, and Recursive Feature Elimination.
Made with: Python, Pandas, Scikit-Learn, Plotly
Personal Website
I created a personal website with HTML, CSS, and JavaScript to further demonstrate and expand upon my projects and experiences. Users can interact with the website through the various clickable buttons and links. The website is dynamic, so shrinking and expanding the site will cause the contents to adjust accordingly. Website is hosted via GitHub.
Made with: HTML, CSS, JavaScript, GitHub
