30 min technical talk aimed to researchers from Carnegie Mellon University from DDMLab. I presented the technical challenged from algorithmic challenges in data-centric interpretable approahces, Graph Fairness and long-term effects and feedback loops.
The European Parliament sent to ELLIS Alicante their JURI Committee on Legal Affairs, which are responsables of the AI regulation in Europe (AI Act). I presented the current Challenges of Algorithmic discrimiation in AI, getting close technical, social and legal concepts in the same presentation. We then discussed the current gaps between technicians and regulators to overcome the current limitations of AI regulation.
A high level introduction about ethics in Artificial Intelligence, why does discrimination occurs and how to solve it using AI [Scientific Dissemination talk].
A tutorial on Graph Rewiring using learnable spectral node embeddings and distances. We first explain the introductions to spectral theory, then we go trought the proposed methods for transductive rewiring, then we delve into how to performe inductive rewiring in graphs using DiffWire and we finally explain the implications of rewiring in Algorithmic Fairness [Accepted Academic Tutorial].
A high level introduction about algortihmic discriminations and lack of diversity and the AI challenges to solve this problem [Scientific Dissemination talk].