Deep Learning to predict patient drug experience
Using online medication reviews, we carried out a sentiment analysis to predict overall patient satisfaction with various Neural Networks.
Our models take unstructured text as an input and output a sentiment score and achieve levels of accuracy in line with current research.
Angelo Cozzubo and Noah Selman
Narrating the Economic Crisis in Lebanon with Data Viz
Using different visualization tools (Altair, Vega, D3) I created a suite of static and interactive graphs in the form of a slideshow to
tell the story of the economic and ensuing food insecurity crisis in Lebanon from 2019 until today.
Leveraging Big Data to Build a Transportation Web App
For my Computer Science elective course, I built an app displaying Chicago transportation statistics using over 20 million rows
of Divvy (bike share), CTA rail and bus ridership, and weather data. I put to use a number of Big Data tools and programming languages
to create functioning batch, serving, and speed layers and a couple of ML models.
Predicting Covid-19 Cases and Deaths with Machine Learning
Using data on government policy measures and national health and economic indicators, we created a tool to
predict Covid-19 related cases and deaths across different countries in the world.
Collaborated with: Diego Diaz and Piyush Tank
Peruvian Party Switching:
Ideologists or Opportunists?
We analyzed and characterized tendencies in party switching among 2020 parliamentary candidates.
Tasks carried out include data viz and designing Selenium-based crawler.
Collaborated with: Andrei Bartra, Angelo Cozzubo, and Marc Richardson