Deep Learning approaches to Goal Recognition

Lorenzo Serina (Information Engineering, Università degli Studi di Brescia)

Abstract

The Goal Recognition task is a problem in the field of Artificial Intelligence, with significant applications across different domains. The current state-of-the-art approach involves transforming the Goal Recognition problem into one that can be solved using classical planning algorithms. This work aims to extend the boundaries of existing state-of-the-art systems by applying deep learning models to the Goal Recognition task.