AI-based decision-making has already penetrated into almost all spheres of human life, from content recommendation and healthcare to predictive policing and autonomous driving, deeply affecting everyone, anywhere, anytime. Such decision models result from complex, multi-step, and interconnected pipelines that orchestrate steps like raw data selection and integration, data preprocessing, model selection, model evaluation, etc. Each step requires decisions, for example, which data to choose, what preprocessing to apply and how to evaluate. These decisions typically made by data scientists may introduce biases and errors in the process resulting in underperforming models and/or models that discriminate against individuals or groups of people based on protected attributes like gender or race. The area of responsible AI has recently emerged in an attempt to put humans at the center of AI-based systems by considering aspects, such as fairness, reliability, and privacy of AI systems. Still, most of these works focus on model creation and, typically, ignore the performance of these models in the long run, i.e., after deployment in real life.
In the first part of the talk, we will discuss responsibility aspects, namely fairness and explainability, during model creation. In the second part, we will cover responsibility aspects after model deployment, focusing on the effect of distribution shifts.
-
Eirini Ntoutsi is a professor for Artificial Intelligence at the Free University (FU) Berlin, since March 2021. Prior to that she was an associate professor of Intelligent Systems at the Leibniz University of Hanover (LUH), Germany. She is a member of the L3S Research Center. Prior to joining LUH, she was a post-doctoral researcher at the Ludwig-Maximilians-University (LMU) in Munich, Germany in the group of Prof. H.-P. Kriegel. She joined the group as a post-doctoral fellow of the Alexander von Humboldt Foundation. She holds a PhD in Data Mining from the University of Piraeus, Greece under the supervision of Prof. Y. Theodorodis and a master and diploma in Computer Engineering and Informatics from the University of Patras, Greece.
Her research interests lie in the fields of Artificial Intelligence (AI) and Machine Learning (ML) and can be described as designing intelligent algorithms that learn from data continuously following the cumulative nature of human learning, while ensuring that what has been learned helps driving positive societal impact.
The lecture will take place online. You can attend the lecture with the following link:
https://hhu.webex.com/hhu/j.php?MTID=m5ab2e51ead58dddc9aafd1a003a1875e
