CSE 163

Algorithmic Fairness and Ethics

​

Arpan Kapoor

Summer 2026��💭Icebreaker (discuss with neighbors):

How are you going to celebrate finishing the quarter?

Add to our Slido!

​

​

slido.com

#2348882

Announcements

• Project Part 3 due Today at 11:59pm on Gradescope!
• Please reach out if you need additional time.
• Final Exam is this Friday, at 1:10 PM in this room!
• We will begin promptly at 1:10 PM, be in your seat by this time.
• Resubmission Cycle 6 closes this Friday at 11:59 PM
• Please reach out ASAP if you have any grade inconsistencies.
• Please complete the course evaluation with your honest thoughts on the quarter 🙏
• This is extremely important to me as a first-time instructor!
• You can find the link in your student email; search for “Course Evaluation CSE 163”

slido.com

#2348882

Group Fairness

• Intent: Avoid discrimination against a particular group, as to avoid membership in the group negatively impact outcomes for people in that group.
• Does not say which groups to protect, that’s a decision of policy and societal norms
• Can be extended to notions of belonging to multiple identities (e.g. intersectionality), but we focus on protecting a single property at this time.

• Usually defined in terms of the mistakes the system might make.

slido.com

#2348882

Definition of Fairness*

• Equality of False Negatives (equal opportunity): False negative rate should be similar across groups.

* Many others exist, many are in form of equations on this confusion matrix. There are other notions of fairness too!

College admission example: P = Successful in college, N = Not successful in college

slido.com

#2348882

Definition of Fairness*

• Equality of False Positives (predictive equality): False positive rate should be similar across groups.

* Many others exist, many are in form of equations on this confusion matrix. There are other notions of fairness too!

College admission example: P = Successful in college, N = Not successful in college

slido.com

#2348882

Human Choice

• There is no one “right” definition of fairness. They are all valid and are simply statements of what you believe fairness means in your system.

• It’s possible for definitions of fairness to contradict each other, so it’s important that you pick the one that reflects your values.

• Emphasizes the role of people in the process of fixing bias in ML algorithms
• Algorithms will do their best to optimize what is asked of them
• This may lead to unforeseen consequences or behavior

slido.com

#2348882

Tradeoff between Fairness and Accuracy

• We can’t get fairness for free, generally finding a more fair model will yield to one that is less accurate.
• Intuition: We saw lots of examples where bias was a byproduct of an “accurate” model since that model was not trained with fairness in mind.

slido.com

#2348882

Pareto Frontiers

slido.com

#2348882

Discussion Guidelines

• Open this slide deck

• You will be working with a discussion group and at least one of the case studies covered in the lesson.
• Introduce yourself to your groupmate(s)!
• Pick at least one note-taker and one presenter.
• Use the questions in the document as conversation starter, but feel free to come up with other questions to discuss!
• Feel free to move around!
• We will repeat this process for the Ethics case studies once we finish discussing the Fairness ones!

slido.com

#2348882