AI-with-the-best Fairness Presentation

This document discusses fairness in machine learning. It notes that as AI touches more aspects of everyday life, fairness becomes more important. However, algorithms can reflect and even amplify biases from their training data. The document explores examples of biases in word embeddings and computer vision models. It also discusses challenges in defining and achieving fairness, as different definitions of fairness can conflict. While machine learning holds promise for reducing biases, developing truly fair systems will require open-minded and ongoing work.

Uploaded by

Virgo Starlight

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

59 views71 pages

AI-with-the-best Fairness Presentation

Uploaded by

Virgo Starlight

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 71

Fairness in Machine Learning

Fernanda Viégas @viegasf

Martin Wattenberg @wattenberg
Google Brain
As AI touches high-stakes aspects of everyday life,
fairness becomes more important
How can an algorithm even be unfair?
Aren't algorithms beautiful neutral pieces of mathematics?

The Euclidean algorithm (first discovered in 300

BCE) as described in Geometry, plane, solid and
spherical, Pierce Morton, 1847.
"Classic" non-ML problem: implicit cultural assumptions
Example: names are complex.

...
Brilliant, fun article.
Read it! :)

Patrick McKenzie
http://www.kalzumeus.com/2010/06/17/
falsehoods-programmers-believe-about-na
mes/
What's different with machine learning?
Algorithm, 300 BCE

Classical algorithms don’t rely on data

What's different with machine learning?
Algorithm, 300 BCE

Classical algorithms don’t rely on data

Algorithm, 2017 CE

ML systems rely on real-world data and

can pick up biases from data
Sometimes bias starts before an algorithm ever runs…
It can start with the data
Sometimes bias starts before an algorithm ever runs…
It can start with the data

A real-world example
Can you spot the bias?
Can you spot the bias?
Model can’t recognize mugs
with handle facing left
How can this lead to unfairness?
word embeddings
Word embeddings

Distributed Representations of Words

and Phrases and their Compositionality
Mikolov et al. 2013
Meaningful directions
(word2vec)
Can we "de-bias" embeddings?
Can we "de-bias" embeddings?
Bolukbasi et al.: this may be possible.

Idea: "collapse" dimensions corresponding

to key attributes, such as gender.
How can we build systems that are fair?

First, we need to decide what we mean by “fair”...

Interesting fact:
You can't always get what you want in terms of “fairness”!
Fairness: you can't always get what you want!

COMPAS (from company called Northpointe)

● Estimates chances a defendant will be re-arrested
○ Issue: "rearrest" != "committed crime"
● Meant to be used for bail decisions
○ Issue: also used for sentencing

https://www.propublica.org/article/machine-bias-risk-assessments-in-criminal-sentencing
This conclusion came from applying COMPAS to historical arrest records.
https://www.propublica.org/article/how-we-analyzed-the-compas-recidivism-algorithm
Enter the computer scientists...
Low-risk
FAIR
Low-risk
FAIR

Medium-high
FAIR
Low-risk
FAIR

Medium-high
FAIR

Did not reoffend

UNFAIR

Black defendants who did not reoffend were

more often labeled "high risk"
Unless classifier is
perfect, can't all be fair
due to different base
rates

Low-risk
FAIR

Medium-high
FAIR

Did not reoffend

UNFAIR
https://www.propublica.org/article/bias-in-criminal-risk-scores-is-mathematically-inevitable-researchers-say
Lessons learned

● Fairness of an algorithm depends in part on how it's used

● In fairness, you can't always get (everything) you want
○ Must make a careful choice of quantitative metrics
○ This involves case-by-case policy decisions
○ These tradeoffs affect human decisions too!
● Improvements to fairness may come with their own costs to other values we want to
retain (privacy, performance, etc.)
Can computer scientists do anything besides depress us?
Hardt, Price, Srebro (2016)
On forcing a threshold classifier to be "fair" by various
definitions:

● Group-unaware
Same threshold for each group

● Demographic Parity
Same proportion of positive classifications

● "Equal opportunity"
Same proportion of true positives
Can computer scientists do anything besides depress us?

A visual explanation...
Attacking discrimination in ML mathematically