Skip to content
View claireboyd's full-sized avatar
19 followers · 11 following

Achievements

Achievement: Pull Sharkx3Achievement: QuickdrawAchievement: YOLO

Achievements

Achievement: Pull Sharkx3Achievement: QuickdrawAchievement: YOLO

Block or report claireboyd

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
claireboyd/README.md

Hi there 👋

My name is Claire Boyd and I am currently a data scientist with NYC's Department of Finance Property Modeling team, where we assess the value of all residential and commercial properties in the city annually. I graduated in 2024 from UChicago's Masters in Computational Analysis and Public Policy program. Before that, I was a researcher at the Urban Institute for 4 years, focusing on racial equity analysis and equitable grantmaking practices for foundations and federal funders.

Claire's GitHub stats

Feel free to explore the following projects/coursework below to get a better sense of my skills and interests:

  • 🔧 Open Source Tool Onboarding @ NYC (Ongoing): Created a series of guides that document the set-up of city machines for open source tools. These hopefully should all be useful for any city employee but they were designed for onboarding new data scientists or analysts at the property modeling team at the NYC Department of Finance (DOF)

  • 📚 Interactive Data Dictionary (Ongoing): Built an interactive data dictionary to make it easier for our team to explore the data we have currently available across our multiple data assets. This can be updated directly from the database metadata, and is supplemented with our own legacy documentation. Helpful when trying to locate a random field across tables or seeing if we collect a certain type of property characteristic.

  • 🍽️ Dirty Comments, Clean Plates (March 2024): Used a corpus of text-based Yelp restaurant reviews to train a model to classify if a restaurant is likely to fail a health inspection and predict if a review is human-generated or generated by OpenAI's GPT 3.5 or 4.

  • ☎️ 311 Requests in Chicago (December 2023): Created a simple web app which gives users a summary of the 311 requests in their Chicago neighborhood, built with Lambda Architecture principles using Apache's tech stack (HDFS, Hadoop, Hive, Spark, etc). The cluster that the app was built with is no longer active, so watch the video included to see the app in action!

  • 🐀 Predicting Neighborhood-Level Rat Activity in New York City (September 2023): Created a time series predictive model to forecast the volume of weekly rat-related 311 requests in each neighborhood, informing mitigation strategy for NYC Rat Czar.

  • 🏠 Finding comparable properties with LightGBM (June 2023): Developed a new feature, the "comparable finder," into the R package lightsnip which enables accurate identification of comparable properties crucial for assessing property values in Cook County. Read more here for a longer explanation of this project.

  • 📈 COVID-19 Online: How people interacted with government websites during the pandemic (March 2023): Working alongside three classmates, we built a complete data pipeline (from collection to visualization) to explore web traffic to HHS websites during the pandemic.

  • 📝 Publications (Aug 2018-Present): This repository has a list of my recent publications from my time as a researcher at the Urban Institute's Office of Race and Equity Research.

For more information about any of the above, please feel free to explore my current resume, reach out via email or connect with me on LinkedIn. Looking forward to connecting!

Pinned Loading

  1. 311requests_chicago 311requests_chicago Public

    Created a simple web app which gives users a summary of the types of 311 requests in their Chicago neighborhood, built with Lambda Architecture principles using Apache's tech stack

    HiveQL

  2. predicting_rats_nyc predicting_rats_nyc Public

    Used time series modeling of 750K relevant 311 requests to predict the volume of weekly rat activity for every NYC neighborhood, with an average MAPE of 5.6%, which currently feeds into a live Powe…

    Jupyter Notebook 1

  3. silent_p silent_p Public

    Group project for Machine Learning for Public Policy, using LeNet and Logistic Regression for a computer vision binary classification task

    Jupyter Notebook

  4. covid19_online covid19_online Public

    Course project that uses a complete data pipeline (from collection to visualization) contextualizing traffic to HHS websites during the COVID-19 pandemic.

    Python