Hi! I'm
Ethan Shapiro 👋
I'm in 4th and final year studying Data Science at UC San Diego with a passion for machine learning applications in gaming and esports 🎮. I also have experience developing using full stack technologies, some of which I used to build this website! Please take a look around and let me know what you think in the contact form 😊
Experience
September 2023 - Present
Meeting weekly with the professor to experiment with adding feature learning to XGBoost.
Researching further how deep neural networks learn features on a variety of computer vision and tabular tasks.
Verified the validity of the Deep Neural Feature Ansatz to understand how deep neural networks learn features.
Tags: Python, Pytorch, Jupyter Notebooks, LateX, Docker, Yaml
Some things I've Made
Featured Project
MOBA Win Prediction
Massive Online Battle Arena (MOBA) games are some of my favorite. What makes them so fun is the teamwork and strategy that comes into play. My friends and I wanted to see how accurately we could predict whether we would win and so that's why we created this project.
Python
Sci-kit Learn
Pandas
Research Paper
Featured Project
Graph Database Recommendation
With over 160 different characters each with their own unique abilities and playstyles, it's hard to know who you might enjoy playing next. This tool I created over a two week period as my internship application to Tigergraph is just the beginning of potentially using graph databases for recommendations.
Python
Graph Databases
Plotly
Heroku
Other Noteworthy Projects
Working in a team, we aimed to predict the outcome of a NYPD complaint at the time it's recieved. In order to do so, we cleaned the data, performed feature extraction and anaylsis, tested multiple models, and did a fairness analysis.
Python
Data Structures
Auto-balancing Tree
Working in a team, we aimed to predict the outcome of a NYPD complaint at the time it's recieved. In order to do so, we cleaned the data, performed feature extraction and anaylsis, tested multiple models, and did a fairness analysis.
Python
Matplotlib
Sci-kit Learn
Pipelines
A minimal web app for managing computer spaces. View how long someone has spent at a computer, queue players when computers are full, and see which computers will open up next.
Assessment of Missingness
Hypothesis Testing
Permutation Testing
This is an ongoing mini project that I use to practice writing Data Structures and various Algorithms.
C
C++
Data Structures
Sorting Algorithms
Working on a team of 4 mentored by Deloitte Data Scientists, we did a comprehensive anaylsis of the impact on job performance and health from COVID-19.
Github Pages
Data Mining
Tableau
Future Forecasting
Created two Convolution Neural Networks with one using transfer learning to classify the MNIST digits datset with 99% accuracy and CIFAR-10 images with ~70% accuracy.