Pozvánka na přednášku semináře

  • David M. Cerna (Department of Artificial Intelligence, ICS Prague):

    Learning Logic Programs with negation, Predicate Invention, and Higher-Order. Definitions Through the learning from Failures Paradigm.

    07.02.2023 14:00Místnost 318 (zoom) @ Ústav informatiky
    Pod Vodárenskou věží 2
    Praha, 182 00
    Hora Informaticae

    Inductive Logic Programming (ILP) is a form of symbolic machine learning that learns clausal theories to explain sets of positive and negative evidence. Learning complex clausal theories that generalize well remains a formidable challenge. Extending the language over which the clausal theories are constructed results, not only in improved generalization, but also the ability to solve tasks that cannot be properly formulated in the simpler environment. Problematically, such language Extensions lead to unsoundness of existing approaches. In this talk, we will introducing the learning from failures paradigm (developed by Andrew Cropper and Rolf Morel) and described extension of the approach that soundly handle the addition of negation, Predicate Invention, and Higher-Order Definitions. (Joint work with Stanisław J. Purgał, Cezary Kaliszyk, and Andrew Cropper).

Předchozí přednášky