Developed an image-to-image translation application with the ability to learn the mapping between input and output images using a semi-supervised learning method.

Utilized the Generative Adversarial Network (GAN) model which consisted of a generator model and a discriminator model – the generator constructed images with a random latent vector space as input, and the discriminator was used to distinguish between authentic images stored in a database and the images constructed by the generator, classifying the images as either “Real” or “Fake.”

Developed an image classifier based on a supervised learning model to categorize and label groups of pixels within an image based on patterns and textural characteristics.  

Optimized validation accuracy by performing a comparison evaluation on multiple models to determine the ideal model for classifying images, which was further enhanced with data augmentation.

Developed an interactive dashboard utilizing the JavaScript D3.js library to display visualizations for users to evaluate and compare stocks allowing for more informed investment decisions.

Obtained stock market data from public data sources that extended over a one-year period, which was cleaned and manipulated for ingestion by the dashboard.

 Performed a statistical analysis of geographical and historical data relating to the locations, magnitudes, and types of natural disasters that resulted in property damage throughout the United States to calculate an overall risk score that could be referenced by home buyers, real estate developers, and insurers to make data-driven investment decisions

 Collaborated with a team of analysts and engineers to identify, collect, and clean relevant data from a variety of public sources utilizing a variety of advanced ETL techniques and open-source tools

 Compiled the results of our analysis and utilized an interactive data visualization tool (Tableau) to summarize our findings

Worked with five team members to develop an interactive web application designed to display political statistics on voters and elected officials to help these groups understand similarities, areas of interest, and other priorities.

Designed all front-end features of the application with interactive visualizations connected to a SQLite database using Python, Django, HTML, CSS, JSON and JQuery.

Presented final project before peers, evaluators, and stakeholders, demonstrating features and functionality