Towards a more inclusive world: Uncovering and addressing human and algorithmic limitations

  • * Register (or log in) to the Neural Network to add this session to your agenda or watch the replay

  • Date
    7 April 2025
    Timeframe
    17:15 - 18:15 CEST Geneva
    Duration
    60 minutes
    Share this session

    Dr. Oliver will first briefly present ELLIS Alicante (https://ellisalicante.org), the only ELLIS unit that has been created from scratch as a non-profit research foundation devoted to advancing AI for Social Good. She will then present a few examples of the work at ELLIS Alicante on uncovering and mitigating both human and algorithmic limitations with AI. 

    On the human front, Dr. Oliver will present her unit’s study of the attractiveness halo effect in the era of AI-based beauty filters, and the research that they have done about the influence of beauty filters that are so popular on social media. On the algorithmic front, she will briefly present FairShap, a novel pre-processing approach based on Shapley Values, along with recent work on measuring structural patterns in social networks and on uncovering the uneven impact of Deep Ensembles. 

  • Are you sure you want to remove this speaker?