Course Details
Country:
Netherlands
Institution:
Free University of Amsterdam
Course Title:
Fair, Transparent and Interpretable Machine Learning
Course Number:
E_EOR3_FML
Course Description:
Machine learning algorithms are increasingly used to make or improve predictions, which then serve as a basis for decision making. Examples include bank lending, college admissions, and bail decisions in criminal proceedings. Though the use of algorithmic decision making is often justified as being "more objective" than human decision making, there are many instances demonstrating that it can produce biased or discriminatory predictions or decisions that unfairly disadvantage certain individuals or groups. Awareness of this issue and knowledge about approaches to address it are of high importance for data scientists and policy makers. Another highly relevant aspect for decision making based on data is the interpretability of the estimation outcomes and decisions obtained using a machine learning method. One possibility is to restrict the class of applied algorithms to interpretable models (e.g. decision trees, linear regression, logistic regression).
Language:
English
Approved Equivalent:
INTS 399
Attachment Files:
Studyguide (1)_13.pdf