How do you make an algorithm fair?

This May, 20 students came together for a week long exploration of “fairness” in algorithms.


Bernard Parker, left. Dylan Fuget, right. Photo by Josh Ritchie for ProPublica Credit: ProPublica. Photo credit: Josh Ritchie for ProPublica.

The basis for the exploration was an article in ProPublica that asks whether an algorithm that assesses recidivism risk is applied fairly to all offenders? Read ProPublica's article on Machine Bias.

The goal was to teach students how to do data science research


ProPublica analyzed the COMPAS recidivism algorithm. Read here.


A group of students created a flowchart to visualize ProPublica’s definition of recidivism

How to read the flow chart and understand how recidivism was defined.

Here’s what some students had to say about the article


The bootcamp participants analyzed the raw data in order to gain new insights into factors that might be predictive for recidivism. Here are their analyses.

ProPublica's Model 2.0

Crime Throughout the Year in Broward County, Florida

 

Threat of Recidivism

 

Data Science Bootcampers

Data Science Scholars Program Students
    • Arpita Shah
    • Major: Statistics, master’s student
    • Project: Data Science Curriculum Development
    • Professors Tian Zheng and Gabriel Young, Statistics
    • Jenna Schoen
    • Major: Department of English and Comparative Literature, doctoral student
    • Project: Training an Optical Character Recognition System on a Corpus of Medieval Manuscripts
    • Professor Dennis Tenen, Department of English and Comparative Literature
    • Young Cho Lee
    • Major: Statistics, master’s student
    • Project: Real-Time Brain State Classification
    • Professor Xiaofu He, Psychiatry (CUMC)
    • Chin-Wen Chang
    • Major: Data Science, master’s student
    • Project: Building Computational Tools to Investigate the Relationship between Microbial Communities and Patient Outcomes
    • Professor: Itsik Pe'er, Computer Science

Data Science Scholars Program Students
    • Ming Li
    • Major: Statistics, master’s student
    • Project: Insights Into Asset Management From Big Data and Natural Language Processing
    • Professor Abis Simona, Business School
    • Gianmarco Saretto
    • Major: Department of English and Comparative Literature, doctoral student
    • Project: Training an Optical Character Recognition System on a Corpus of Medieval Manuscripts
    • Professor Dennis Tenen, Department of English and Comparative Literature
    • Abhi Gupta
    • Major: Computer Science, sophomore
    • Project: Visual-Tactile Geometric Reasoning
    • Professor Peter Allen, Computer Science
    • Nav Ravindranath
    • Major: Applied Mathematics, master’s student
    • Project: Human Peri-Ictal Single Unit Activity
    • Professor Catherine Schevon, Neurology (CUMC)

Data Science Scholars Program Students
    • Tingyu Mao
    • Major: Electrical Engineering, master’s student
    • Project: Visualization of Continuous Health Data Measurements
    • Professor: Samuel Sia, Biomedical Engineering
    • Cindy Xu
    • Major: Statistics, 2018, master’s student
    • Project: Informatics Approaches to Biomedical Evidence Appraisal Using Public Data
    • Professor Chunhua Weng, Biomedical Informatics (CUMC)
    • Jonathan Shor
    • Major: Computer Science, master’s student
    • Project: Neuronal Ensemble Detection with Temporal Conditional Random Field
    • Professor Rafael Yuste, Biological Sciences
    • Maxwell Bohn
    • Major: Mathematics and Statistics, junior
    • Project: Analysis of Loan Data From Fannie Mae and Freddie Mac
    • Professor David Rios, Department of Statistics

Data Science Scholars Program Students
    • Vivian Casillas
    • Major: Political Science & Statistics, senior
    • Project: How Latin American Countries Contend with Anti-Competition
    • Professor Sharyn O'Halloran, SIPA
    • Robbie Netzorg
    • Major: Computer Science and Math, senior
    • Project: Twitch Popularity Project
    • Professor Eugene Wu, Computer Science
    • Lauren Arnett
    • Major: Computer Science, senior
    • Project: Twitch Popularity Project
    • Professor Eugene Wu, Computer Science

Data Science Scholars Program Students
    • Gary Buranasampatanon
    • Major: Data Science, master’s student
    • Projects: 1. Examine How Investors Evaluate Soft Information and 2. Evaluating Evolution of Economists’ Tools
    • Professors Simona Abis and Anton Lines, Business School
    • Mary Liu
    • Major: Applied Mathematics and Statistics, senior
    • Project: Climate and Physics
    • Keshan Huang
    • Major: Economics-Statistics and Astrophysics, sophomore
    • Project: Technology Transfers From the University to the Marketplace
    • Professor Orin Herskowitz, Technology Ventures
    • Michelle Yang
    • Major: Electrical Engineering and Computer Science, UC Berkeley 2018 graduate
    • Project: Developing a System Called the Deep Neural Inspector
    • Professor Eugene Wu, Computer Science


550 W. 120th St., Northwest Corner 1401, New York, NY 10027    212-854-5660
©2018 Columbia University