Algebraic and Logical Methods for Explainable Artificial Intelligence
(ALADDIN)
The project aims at developing mathematical methods for explainable artificial intelligence.
The project is funded by the Science Fund of the Republic of Serbia (Science Fund)
Serbian Science and Diaspora Collaboration Program: Knowledge Exchange Vouchers
The Research field of the Project: Mathematics and Computer Science
Mission of the ALADDIN project
The project aims at developing mathematical methods for Explainable Artificial Intelligence (XAI). The emerging field of XAI seeks fundamental approaches that contribute to the interpretability and explainability of deep neural networks (DNNs) as opposed to practices where DNN is commonly used as an opaque box. The focus of the project is on specifying algebraic structures and formal techniques for performing analysis of deep learning models' behavior in biomedical image processing.
Effort
To establish research collaboration focused on the identification and development of fundamental algebraic structures and formal techniques underlying and application of artificial intelligence in image processing.