Past Projects
Sign Language Prediction Program - Python
The program utilizes machine learning techniques to compare and contrast three classification methods: Decision Tree, K-NN, and Convolution Net. The dataset was obtained through Kaggle.com, where we chose to assess the accuracy of a trained model's ability to identify hand signs given American Sign Language correctly.
Convolution Net Accuracy:
The graph depicts the training and validation accuracy after over 5000 iterations of training utilizing the convolution net model to assess the similarity of pixel colours a given hand sign in comparison to the trained data.
Music Player Website - JavaScript, Numpy, Vue, NoSQL
A website where users create an account to access a library of music, to then create playlists of their favourite songs. The website utilizes Node.js in addition to MongoDB to store user’s information such as login credentials and playlists. Users can compile songs into a favourited playlist and listen to them at any time. Users can also take quizzes based on genre or artist, trivia stats are saved upon completion in a JSON file.
Physics-Based Plinko Simulation - Python
Utilizing Pygame, Numpy, and Scipy to create a working simulation of the minigame “Plinko” from the famous gameshow “The Price is Right”. Includes the base simulation alongside a Monte Carlo demonstration to introduce the topic of randomness as a contestant “releases” the puck, akin to the real show.
Social Music App (Aud.io) - Flutter, SQL
A mobile application that was tested through Android Studio. Allows users to log in to a custom profile, where they can then add favorited music genres. They can then search for all global users and find those with similar tastes in music. The application also allows you to view the most liked genres statistically across all users. Users can also swap languages between English, Spanish, and French.
