I am currently a Senior at University of Memphis majoring in Computer Science.
In my free time I love to work on a multitude of different projects, whether it be something that involves Machine Learning or maybe even a video game!
Scripted in Python to build a bot with the ability to post, follow, and like asynchronously.
I used a SQLite database as the backend database to store information/state and wrote SQL to query the database.
Utilized multitude of Python libraries and APIs such as PRAW, Beautiful Soup, Requests, and RegEx to scrape media from Reddit. Furthermore, it able to maintain itself by removing images older then a date defined in the config file.
Source code
This video game was created as part of Ludum Dare 42! I worked with a team consisting of artists and sound engineers, but I did all the game development/coding through Unity and C#. It deploys procedural generation by utilizing a few different algorithms.
The Algorithms used:
Built from Donkey Car kit and API. Originally the API uses behavior cloning, but I decided to implement my own behavior through the usage of TensorFlow. Tha car is able to detect certain object, such as a stop sign, and stop accordingly. Also, modified the behavior to prevent collisions through using an Ultrasonic Sensor.
Video of it!
Source code
A language model built with Keras, using TensorFlow as the back-end, and Python. There is also an accompanying demo webpage using HTML, CSS, and JavaScript. The model will predict the next word given a sequence of words. Additionally there is an option to use Locality-sensitive hashing (LSH) to find similar users to you and recommend words using each similar user's language model. In the repository I have a report on how it works (docs/Report/Word_Suggestions_Report.pdf).
Source code
GrantFinder is a Web App built with Ruby on Rails. Its purpose is to allow researchers to easily find grants from the National Institutes of Health (NIH) website by using the Search Engine I built. Initially I scrapped descriptions of each grant from the website to populate the database. Next, I applied Latent Semantic Analysis (LSA) on the dataset to create a semantic model. Finally, when a user enters in a query, it will compute the cosine similarity to find related grants.
Source code
A video game built using Unity, C#, and numerous APIs. The objective of the game is try to stop chickens from crossing the road. I built the entire game myself including the sprites and sounds! It is currently on Google Play and the App Store!
Close Project
An item-item collaborative filtering algorithm that is trained on a subset of the Netflix dataset. Given a valid user ID and movie ID it will use k-nearest neighbors algorithm to predict the rating that the user would give the movie. I built everything from scratch so no dependencies required!
Source code
A video game created using Unity and C# for Global Game Jam 2019. Play as a hermit crab trying to find a home while battling other hermit crabs for shells! Worked/managed at team that included an artist, sound engineer, and another game developer.
Play the game!
The Witching Hour is a short horror game about a malevolent spirit that inhabits the house you've just inherited. Possessed appliances are antagonizing the being by transmitting spectral sounds. Keep the appliances quiet to stay alive! I helped as a programmer for the video game through using Unity and C#. The core game was developed in 48 hours for Global Game Jam 2018 with the theme Transmission. Worked with a team of many different types of people: programmers, music composer, game designers, 3D artist, and more!
Play the game!
A car customization demo I made while interning at AutoZone. It was built using React, JavaScript, HTML, and CSS. Also, had to use Blender for the car model and WebGL to render it on the screen.
Demo Video
Source code
Built Huffman Coding Compression Algorithm completely using Ruby. It is able to encode and decode a file.
Source code
The purpose of this project was to create a reliable data transfer over UDP's unreliable service. So, to enable this, I used two different implementations: Selective Repeat and Alternating-bit. The client.py script will request a file from server.py. In the event that server.py does get passed a valid file, it will then begin to transfer over the file using the Application layer reliable data transfer. This was built using Python and Python socket module.
Source code