Generative Adversarial Networks (GANs) are powerful machine learning tools designed to generate data. The goal of our package was to create a framework to simplify the development of GANs, as they are often complex in nature. We created a Python package with object-oriented representations of GANs, where a user could create and train their own GAN. We designed a tutorial to explain how to use our package, tested our tutorial on developers, and updated our tutorial based on their feedback. We found that users were successful in developing GANs by following the tutorial and using our package.

• This report represents the work of one or more WPI undergraduate students submitted to the faculty as evidence of completion of a degree requirement. WPI routinely publishes these reports on its website without editorial or peer review.

Creator

• Costello, Kyle
• Astor, Ryan
• Lewis, Alek

Subject

• Innovation
• Computing
• Learning
• Project-Based Learning

Publisher

• Worcester Polytechnic Institute

Identifier

• 53076
• E-project-032322-141257

Keyword

• Package
• Python
• Machine Learning
• GAN
• Simplify
• Development

Advisor

• Rundensteiner, Elke A.

Year

• 2022

UN Sustainable Development Goals

• 9 - Industry, Innovation and Infrastructure

Date created

• 2022-03-23

Resource type

• Major Qualifying Project

Major

• Computer Science
• Data Science

Rights statement

• In Copyright