Login for PhD students at UCPH      Login for others
Algorithmic Fairness and Bias in Machine Learning (Online)
Provider: Faculty of Health and Medical Sciences

Activity no.: 3168-21-00-00 
Enrollment deadline: 08/02/2021
Date and time08.03.2021, at: 09:00 - 11.03.2021, at: 16:00
Regular seats20
Course fee4,080.00 kr.
LecturersTibor V Varga
ECTS credits2.80
Contact personKathe Jensen    E-mail address: kje@sund.ku.dk
Enrolment Handling/Course OrganiserPhD administration     E-mail address: phdkursus@sund.ku.dk

Aim and content
This course is free of charge for PhD students at Danish universities (except Copenhagen Business School), and for PhD students at graduate schools in the other Nordic countries. All other participants must pay the course fee.

Anyone can apply for the course, but if you are not a PhD student at a Danish university, you will be placed on the waiting list until enrollment deadline. This also applies to PhD students from Nordic countries. After the enrollment deadline, available seats will be allocated to applicants on the waiting list.


Learning objectives
A student who has met the objectives of the course will be able to:

1. Understand how unfair algorithms have the ability to have tangible impact on human lives.
2. Critically think about concepts of fairness and articulate their own definition of fairness.
3. Understand the main aims of machine learning and the most important metrics of predictive modelling.
4. Understand the differences between equality, equity, justice, and understand the concepts of discrimination, disparate treatment and disparate impact.
5. Understand the differences between group fairness and individual fairness.
6. Critically think about various sources of data biases.
7. Align fairness concepts with the main aims of machine learning.
8. Have an overview of solutions to counteract data biases using various methods.
9. Navigate the growing literature of algorithmic fairness and available software packages.

+1: An important objective is to provide inspiration to students from real-life examples and applications from a world of algorithmic fairness.


Content
Algorithms will soon completely surround us - from college applications, bank loans and the criminal justice system, through self-driving cars to clinical decision making, algorithms are becoming an organic part of our lives. In entering this era, we rely heavily on machine learning (ML) algorithms. In many applications, ML algorithms were designed with the intent of removing preexisting human biases and generate fair decisions using mathematical rules, avoiding biased human judgement. Unfortunately though, societal inequities and biases from various other sources taint the historic data that these algorithms use, further propagating biases into decisions for the future. In addition, while most algorithms are tuned to generate the most accurate predictions, less attention is paid to generate fair and unbiased results.

This course introduces the fundamentals of algorithmic fairness, introduces the most common biases in machine learning, discusses recent examples from the algorithmic fairness literature and gives an overview of the most common group- and individual fairness metrics. The course will include lectures from the course director and international expert external speakers, and practical exercises that will help students critically assess the biases in data and in the results of machine learning algorithms.


Participants
PhD students who have interest in algorithmic fairness. The maximum number of participants is 20.
In case the course is forced into an online setting (e.g. due to COVID-19), the lectures can be attended by others, but the practical exercises are always limited to the max. 20 signed up participants.


Relevance to graduate programmes
The course is relevant to PhD students from the following graduate programmes at the Graduate School of Health and Medical Sciences, UCPH:

The course will be relevant for a large number of graduate programmes due to the importance of fairness and machine learning to almost every subsets of medical research.

Public Health and Epidemiology
Biostatistics and Bioinformatics
Clinical Research
Cardiovascular Research
Cellular and Genetic Medicine
Medicine, Culture and Society
Clinical Cancer Research
Basic Metabolic Research


Language
The course will be held in English.
The R programming language will be used for exercises.


Form
A mix of lectures and practical exercises.


Course director
Tibor V. Varga (director)
Assistant Professor
Section of Epidemiology, Department of Public Health, University of Copenhagen
Email: tibor.varga@sund.ku.dk

Naja Hulvej Rod (co-director)
Professor, Head of Section
Section of Epidemiology, Department of Public Health, University of Copenhagen
Email: nahuro@sund.ku.dk


Teachers
When considering external teachers, special attention was paid to gender equity.

Faculty Teacher
Tibor V. Varga, Assistant Professor – University of Copenhagen, Denmark

External Teachers
Thore Husfeldt, Professor – Department of Computer Science, IT University of Copenhagen, Denmark
Kim Bernet, UX and UI Designer & Thérèse Mannheimer, CEO – Grace Health, Sweden
Katarzyna Wac, Professor & Sofia Laghouile, PhD Student – Human-Centred Computing, Datalogisk Institut, University of Copenhagen, Denmark
Timo Minssen, Professor & Audrey Lebret, Postdoc - JUR Centre for Advanced Studies in Biomedical Innovation Law, University of Copenhagen, Denmark
Ricardo Silva, Associate Professor – Department of Statistical Science, University College London, UK


Dates
8 March 2021 (Monday)
- “Introduction to Fairness”, time: 09:00-12:00 (3 hours)
- “Primer in Algorithms and Machine Learning”, time: 13:00-16:00 (3 hours)

9 March 2021 (Tuesday)
- “Metrics of Algorithmic Fairness”, time: 09:00-12:00 (3 hours)
- “Machine Learning and Group Fairness – exercise”, time: 13:00-16:00 (3 hours)

10 March 2021 (Wednesday - external lectures day)
- “Calling bullshit in the digital age”, time: 10:00-12:00 (2 hours)
- “Digital design for fairness”, time: 11:00-12:00 (1 hour)
- “Bias and Discrimination from an algorithmic perspective”, time: 13:00-14:00 (1 hour)
- “Artificial Intelligence Ethics and Human Rights”, time: 14:00-15:00 (1 hour)
- “Counterfactual Fairness”, time: 15:00-16:00 (1 hour)

11 March 2021 (Thursday)
- “Practical Exercise for Fairness – exercise”, time: 09:00-12:00 (3 hours)
- “Practical Exercise for Fairness – exercise”, time: 13:00-15:00 (3 hours)
- “Closing remarks”, time: 15:00-16:00 (1 hour)


Course location
Online


Registration
Please register before 8 February 2021

Seats to PhD students from other Danish universities will be allocated on a first-come, first-served basis and according to the applicable rules.
Applications from other participants will be considered after the last day of enrolment.

Note: All applicants are asked to submit invoice details in case of no-show, late cancellation or obligation to pay the course fee (typically non-PhD students). If you are a PhD student, your participation in the course must be in agreement with your principal supervisor.

Search
Click the search button to search Courses.


Course calendar
See which courses you can attend and when
JanFebMarApr
MayJunJulAug
SepOctNovDec



New courses
Courses are published regularly. High demand courses are announced in spring and autumn.


Learn which courses are announced on fixed dates