Maria Gargiulo
|
Statistician
HRDAG (Human Rights Data Analysis Group) |
|
Maria Gargiulo is a statistician at the Human Rights Data Analysis Group, where she works on record linkage, population size estimation, and uncertainty representation in a variety of international contexts. Before joining HRDAG, Maria conducted field research on intimate partner violence in Nicaragua. She also served as a Data Science Fellow at the United States Census Bureau. Maria holds a Bachelor of Science in statistics and data science and Spanish from Yale University.
You can find her on Twitter @thegargiulian |
Workshop: Data Processing and Statistical Models to Impute Missing Perpetrator Information
Prerequisite: Basics of binary classification, notion that not all ML algorithms handle missing data gracefully and imputation may be required before model fitting
The Human Rights Data Analysis Group (HRDAG) uses methods from statistics and computer science to help answer questions about mass violence using incomplete and unrepresentative datasets. This talk will present the context in which HRDAG works and how open-source tools are crucial to their analytical projects. A specific example of work imputing missing perpetrator information will then be presented to illustrate the transformation from spreadsheets of information about victims to predictions about probable perpetrators of violence. In the process, we’ll review how topic models, missing data imputation, and multi-label classification were used. Along the way, emphasis will be placed on the four core principles that guide HRDAG’s workflow to ensure that statistics about human rights violations are generated with as much rigor and scientific accuracy as possible
The Human Rights Data Analysis Group (HRDAG) uses methods from statistics and computer science to help answer questions about mass violence using incomplete and unrepresentative datasets. This talk will present the context in which HRDAG works and how open-source tools are crucial to their analytical projects. A specific example of work imputing missing perpetrator information will then be presented to illustrate the transformation from spreadsheets of information about victims to predictions about probable perpetrators of violence. In the process, we’ll review how topic models, missing data imputation, and multi-label classification were used. Along the way, emphasis will be placed on the four core principles that guide HRDAG’s workflow to ensure that statistics about human rights violations are generated with as much rigor and scientific accuracy as possible