Application of recurrent neural networks for goal recognition problems

My research concerns the study and application of Machine Learning techniques to Goal recognition problems. In particular I’m currently working on applying Recurrent Neural Networks to a new type of goal recognition problem where we have fewer information on the actions and predicates. This work combines Recurrent Neural Networks with some heuristics on the predictions in order to optimize our performance.


  • Primary: Alfonso Gerevini
  • Other advisors: Ivan Serina

Short Bio

Mattia Chiari was born in Brescia (Italy) on 14th October 1993. He achieved his Master Degree in Computer Engineering at Università degli Studi di Brescia in March 2019. In his Master Thesis he worked together with IBM Research Lab Dublin on heuristics and compression techniques for Compressed Path Databases. In 2020 he joined the Information Engineering Group supervised by Alfonso E. Gerevini at Università degli Studi di Brescia as Research Fellow.