Measuring Algorithmic Fairness: challenges and solutions for the industry
September 04, 1:45 pm - 5:00 pm (CEST)
Tutorial Description: https://ecai20-maxsat-tutorial.github.io/
Tutorial website: https://fair-ml.github.io/Algorithmic-Fairness/
The tutorial will focus on communicating real-world experience on assessing fairness throughout the machine learning model development life-cycle (all elements also relevant to non-machine learning analytical models). It will cover innovative solutions for measuring and correcting algorithmic bias through hands-on exercises in a range of use-case scenarios from banking to policing.
As machine learning models are employed to inform decisions in an increasingly wide range of human scenarios (e.g. banking, criminal justice, medicine and employment), it is very important to ensure that these decisions are equitable. This tutorial aims at providing the audience with an understanding of the nascent field of algorithmic fairness, by analysing the existing approaches in the literature, and complementing and critiquing them with lessons learned from our experience applying them in real-life situations, both in financial services and government agencies.
Exercises will focus on key lessons learned and will cover how to deal with scenarios such as where highly unbalanced datasets mean the state-of-the-art metrics are not sensitive enough and where the defined state of the art metrics fall short in specific modelling contexts, as well as essentials such as how to choose the best fairness metric to use for a given scenario, how to understand the effect of the repair algorithms and when it makes business sense to use them and the need to understand the source of bias prior any other steps and focus on alternative solutions to the repair algorithms.
Medb Corcoran, Managing Director, Lead of Accenture Labs, Dublin.
Role in tutorial: Overview, business context for fairness and SME on real world implementation challenges in banking
Medb is Managing Director of Accenture Labs in Dublin, one of Accenture’s seven key research hubs around the world – where she leads a team of research scientists that apply emerging technology in the area of AI to help solve problems and create value for clients and society. One of the main focus areas of her team is Explainable AI – which is a key element of being able to ascertain if an AI system is fair.
Medb is Accenture-Turing Research Director with The Alan Turing Institute in the UK – where she sets agenda and oversees the Accenture portfolio of joint research focusing on ‘Innovating for the Responsible use of AI’. As part of this role, she participated in an Accenture Turing Hackathon in 2018 and brought one of the outcomes, a POC on quantifying Algorithmic Fairness, to Accenture the Dock and sponsored a project to develop this into an Algorithmic Fairness Tool that year. She then oversaw the application of the tool with a European Financial Services Client. She oversees the annual joint innovation symposium – a video of the 2019 symposium, which featured the Algorithmic fairness tool can be found here: here She presents regularly both internally and externally on Responsible AI, including algorithmic fairness. linkedIn
Steven Tiell, Sr. Principal, Responsible Innovation, Accenture Labs
Role in tutorial: Ethical AI Overview, business context for fairness and SME on Ethical AI Governance
Steven started Accenture’s journey in data ethics in 2013 while leading foresight research for the firm’s annual Technology Vision. Since that time, his pace of discovery in the field has only accelerated. You can read the first set of papers he published in collaboration with over a dozen others in 2016 at Accenture.com/dataethics. In 2018, he started the Data Ethics Salon Series for practitioners to convene and help each other establish best practices – he has spun the Salon Series out to the Atlantic Council’s GeoTech Center to accelerate the global discovery and publication of best practices. He speaks frequently on topics of data ethics, governance, and related issues, often to large, global audiences and he has published extensively in this field. Steven is the author of many publications including: Universal Principles of Data Ethics: 12 Guidelines for Developing Ethics Codes, Facilitating ethical decisions throughout the data supply chain, Ethical algorithms for “sense and respond” systems, Building Data and AI Ethics Committees. linkedIn website
Laura Alvarez Jubete, Analytics Innovation Assoc. Manager at The Dock, Accenture’s Global Innovation Center
Role in tutorial: Data Scientist to teach technical aspects of the tutorial
Laura is an Analytics Lead at The Dock, Accenture’s Global Innovation Center in Dublin, Ireland. Her main research interest is in the area of ethical AI and algorithmic fairness. Laura has led for the last 1.5 years the design and build of Accenture’s Algorithmic Fairness Tool. She has recently completed a 5-month project with a European Financial Services Client where herself and her team conducted the fairness assessment and bias mitigation of two of their classification models, gaining valuable insights as to the challenges and solutions related to the application of state-of-the-art methods coming from academia in a real-life setting. linkedIn
Andreea Roxana Miu, Analytics Innovation Analyst at The Dock, Accenture’s Global Innovation Center
Role in Tutorial: Data Scientist to teach technical aspects of the tutorial
Andreea has been working closely with a European Financial Institution in the area of Algorithmic Fairness for the last 5 months. The collaboration consisted in implementing algorithmic fairness approaches researched in academia and testing their suitability and applicability in industry settings.
Andreea recently graduated a MSc in Data and Computational Science from University College Dublin, where she focused her research thesis on explainability in machine learning. In 2017 she received a BSc in statistics and economics from the Faculty of Cybernetics, Statistics and Economic Informatics in Bucharest, Romania. linkedIn