This project was part of an overarching initiative to improve the technical standards and automation capabilities within laboratory activities for Georgia Tech's Earth Sciences department.
I've collaborated with 2 professors within the Earth Sciences department, Dr. Meg Grantham and Dr. Erin Griffith, to begin this initative with QuakeCaster.
QuakeCaster is a fully automated earthquake simulation lab setup, designed to re-enact the strike-slip faults associated with tremors from the Earth.
In this project, I predicted the launch and landing probabilities of the first stage of the SpaceX Falcon 9 rocket through machine learning, data visualization, and analytics conducted in Python.
I extracted datasets through an open-source SpaceX REST API and used web scraping with Python's BeautifulSoup library to gather additional historical data on SpaceX launch and landing proceedings.
I also conducted exploratory data analysis through SQL, Matplotlib, and Pandas, extracting relationships between key performance variables like payload mass, orbit type, launch site, and mission status.
I created visualizations using Folium and generated an interactive analytical dashboard of results with Python's Plotly and Dash frameworks.
I trained three different machine learning models (Decision Tree Classifier, K-Nearest Neighbors, and SVM) and analyzed which would be the most accurate at predicting whether a given launch would land the first stage successfully.
And lastly, I generated a sample stakeholder report with important conclusions generated from overall data analytics and visualizations.
In this project, I created a web application using React and Flask that analyzes uploaded nutritional labels to determine a user's allergies and minimizes allergen-related risks.
With Python and the Tesseract OCR API, I successfully extracted data from nutritional label images, achieving a 72% accuracy rate.
I developed a REST API that notifies users about all nutritional ingredients to which they might be allergic.
In this project, I designed a Tic-Tac-Toe game offering three difficulty levels, enabling players to challenge an AI opponent through the command line interface.
For the most challenging level, I implemented the Minimax Algorithm with alpha-beta pruning for optimal performance.