Combining Deep Learning and Knowledge Representation and Reasoning: Overview, Limitations and Solutions for Real-World Applications 

4 October 2022 | Online | 10:00 | Pierangela Bruno (University of Calabria, Italy)

Abstract
In recent years, Deep Learning (DL) techniques have been widely used in several domains, such as healthcare context for supporting medical imaging and computer-aided diagnosis. Such approaches emerged as powerful tools to improve healthcare assistance and they have proved to be greatly promising in extracting meaningful patterns from huge amounts of data. However, DL-based approaches suffer from the lack of proper means for interpreting the choices made by the models and for explicitly including available knowledge to drive the decisions.

In this talk, we provide an overview of the most used Deductive and Inductive approaches, with a specific focus on the ones applied on medical data. Furthermore, we will illustrate recent progress along with some innovative solutions proposed to combine the two approaches in practical scenarios. 

Bio
Pierangela Bruno is a Computer Scientist. She got a master’s degree within a Dual Degree Program: Computer Science (University of Calabria, Italy, 2017) and Software Engineering (University of Applied Science of Upper Austria, Austria, 2017), and received her PhD in Mathematics and Computer Science at University of Calabria (2021). She held a post-doc position at the Department of Mathematics and Computer Science (University of Calabria, Italy, 2021). Currently, she serves as non-tenured Assistant Professor (RtdA) in the same Department. Her main research interests involve Deep Learning-based approaches for the analysis of biomedical images with the aim of providing an automated assessment of pathological conditions, detecting, and segmenting specific elements of medical interest.