Natalie Cygan

CS AI @ Stanford



About

Hello! My name is Natalie Cygan and I am a Junior at Stanford University studying Computer Science with a focus in Systems and Artificial Intelligence. I am interested in building technologies that broadly enable advanced forms of human-computer interaction and make intelligent systems useful to the rest of humanity.

With the increasing impact of computing and automation, I also am an advocate for Computer Science education. At Stanford, I've had the wonderful opportunity to spend time section leading for the introductory Computer Science classes. Additionally, I have been involved with the Stanford Society of Women Engineers (SWE) as an outreach intern and my own local efforts to inspire young girls to pursue STEM.



Projects

Classifying Food Deserts

CS 221 Final Project

Food Deserts are regions of the United States with limited access to affordable and nutritious food. Aimed at helping policy-makers and produce-providers understand where to best focus their efforts to alleviate food insecurity, we built models (Random forest, SVM, Neural Network) to classify areas as food deserts using socioeconomic features and to identify which features are the best predictors of food deserts.

python   pytorch

Learning to Groove: Conditional Melody Generation from Authentic Basslines

CS 236 Final Project

Creating music with an artificial “jam session” that uses two generative models: a bassline model that is trained first and provides rich encodings, and a melody model that conditions generation upon those bass encodings. We also present a novel encoding scheme for representing polyphonic music.

python   pytorch

TerraByte: A Climate Change Data Science Learning Platform

CS 398 Final Project

This project was started out of the CS 398: Computational Education class as a means to provide late-middle school and early-high school students with a platform to learn about data science while engaging with climate change and sustainability related issues.

react   javascript

Facial and Portrait-Aware Artistic Style Transfer

CS 231N Final Project

This was my final project for CS231N: Convolutional Neural Networks for Visual Recognition. We propose and implement a portrait-aware style transfer method that treats the subject foreground and background differently, as well as an additional step to preserve facial features.

python   pytorch   jupyter

Dueling Projectile Launcher Robot

Mechatronics Capstone Project

The 2019 ME210 final project challenge consisted of two arenas, in which team-designed robots navigate, attempting to score points by launching projectiles to knock over towers. I was responsible for designing the robot's software and contributed to the launcher design.

C   Arduino