John Drotos

John Drotos is a senior at Columbia University looking to start his career in software development or data analysis. He has a robust technical skillset in programming and mathematics, which is complemented by his creativity, analytical thinking, and leadership.

Follow me on LinkedIn

John Drotos
Summer 2024

Embedding Matrix Evolution in Vision Transformers

Last summer, I had the opportunity to work at the Graphics Imaging and Light Measurement Lab at Columbia University, under the mentorship of Professor Corey Toler-Franklin. Alongside a graduate student team, I contributed to the development of a novel vision transformer architecture designed for tumor detection in medical slides. My primary responsibility was to conduct an experiment aimed at understanding how transformers learn during the fine-tuning process when trained on our medical dataset. I achieved this by generating and analyzing comparisons of the embedding matrices before and after fine-tuning. My research provided valuable insights into the internal mechanisms of the models, which are often seen as black-box technologies, and greatly enhanced the team's understanding of the models we're improving. I am excited to continue working with the lab this fall.

Technical Skills

  • Utilized the open-source MMDetection framework within a Docker container on AWS to train and test models
  • Visualized and analyzed data using Pandas and Matplotlib

Non-Technical Skills

  • Designed a research poster to present at the Columbia Undergraduate Research Symposium in October 2024
  • Independently learned the open-source framework and trained other lab members on how to use it
Fall 2024

NBA Prediction APP

As a personal project, I am developing a machine learning model to predict the point spreads of regular season NBA games. I collected data from the last 15 years using a public API and used Pandas for extensive feature engineering. Some of the features I engineered include season averages, rolling averages, win streaks, home/road performance metrics, and head-to-head matchup statistics. I have implemented and tested both a Random Forest model and a neural network to make predictions. While my results are still in the improvement phase, I am refining my models to enhance their accuracy.

Technical Skills

  • Extracted and processed 15 years of NBA game data from a public API
  • Performed feature engineering using Pandas to create season and rolling averages, win streaks, and head-to-head statistics
  • Developed and tested machine learning models, including a Random Forest and neural network built with TensorFlow