Course Abstract

Training duration: 90 minutes

Recently, academics as well as policy makers have written many papers, on responsible data science / AI. Moreover, many open-source packages for bias dashboards or tools for `fairness’ have been proposed. This course aims to provide attendees a broad overview as well as the specific technical background to use the available ` fairness’ tools. In addition, a governance framework describing the precise responsibilities of data scientists will be discussed.

DIFFICULTY LEVEL: INTERMEDIATE

Learning Objectives

  • Understand what are Unfair/Unethical through examples of applications

  • Explore Policies & Frameworks for fairness

  • How to use ppen-source approaches and techniques for fairness

  • Understand practical example of fairness in retail banking.

Instructor

Instructor Bio:

Principal Data Scientist; Associate Professor, Econometrics| de Volksbank; Tilburg University

Ramon van den Akker, PhD

Ramon van den Akker works as a data scientist at the AI Center of Excellence and the Risk Modelling departments of de Volksbank, a Dutch retail bank located in Utrecht. He also works, as an associate professor, at the econometrics group of Tilburg University. His research interests cover various fields in data science, machine learning, econometrics and statistics and his research findings have been published in leading journals in econometrics and statistics. Ramon has taught courses in data science, econometrics, life insurance, machine learning, mathematics, probability theory, quantitative finance, and statistics at Tilburg University, Tias business school, Tilburg Professional Learning, the Jheronimus Academy for Data Science (JADS), and the Dutch Actuarial Institute. In his work at de Volksbank, Ramon mainly works on projects related to data-driven innovation, but also on governance aspects like frameworks for responsible AI & data science and the use of techniques for privacy-preserving data analytics.

Course Outline

Module 1: Unfair/Unethical Examples of Applications 

- Examples of data science applications that are considered unfair / unethical 

- Main `driving sources’ behind such applications 

Module 2: Policies & Frameworks 

- Proposed policies and frameworks 

- Upcoming regulations across the paradigm 

Module 3: Open-source approaches and techniques for fairness 

- Overview of the (academic) literature 

- In-depth discussion of the similarities and dissimilarities between different approaches

 - Application of open-source Python packages that provide so-called `bias-dashboards’. 

- Using open-source datasets and packages for demonstration 

- Overview of methods that try to enforce fairness by design. 

Module 4: Framework used by de Volksbank (a Dutch retail bank) 

Background knowledge

  • This course is for current or aspiring Data Scientists, Machine Learning Engineers, AI Product Managers

  • Knowledge of following tools and concepts is useful:

  • Python (Jupyter notebook) and supervised ML concepts.

  • The session focuses on concepts and not on technical implementation.

  • Mathematics for data science will be used in order to provide clear definitions.

Real-world applications

  • Large number of Fortune 500 companies are increasingly adopting AI Fairness policies and frameworks in their data science teams

  • Responsible AI frameworks are of increasing importance in finance and insurance industries

  • Requirement for knowledge and expertise in AI fairness are growing in data science teams to reduce model biases