Ryan Franks

Professional Interests

I do machine-learning research and development for Discover Financial Services. Research areas include:

• Identifying and Explaining Why Models Produce Disparate Impact
  • Determining how predictors contribute to bias (Wasserstein-based fairness interpretability framework for machine learning models )
  • Modifying models to reduce bias (Model-agnostic bias mitigation methods with regressor distribution control for Wasserstein-based fairness metrics )
• Building and Testing Deep-Learning Models
  • Tabular Models
  • Sequence-Based Models
  • Graph-Based Models
• Exploring Process Improvements for Conventional Model Building

Personal Interests

• Algorithmic Art
  • Painting with a rigid body physics engine
  • Applying specially designed transforms to fractals and generating corresponding animations
  • Visualizing number factorization
• Simulations
  • Branch Growth Simulations
  • Rigid Body Physics Simulations
• Animal Welfare and Vegan Food

Unprofessional Interests

• Deep Doggos: Poetry comics generated using public domain dog images, neural network style transfer, and my own poetic genius

Contact Me

I can be reached at ryanfranks1996 AT gmail.com