TRUST Workshop

Recently, biomedical and health informatics has registered important developments concerning the application of Machine Learning (ML) models. The support to clinical decision making is one of these domains. However, the integration of these computational models into clinical practice remains constrained.

The main goal is to provide solutions for machine learning explainability, providing the ability to justify its outcomes, thus effectively supporting physicians in the use of the model prediction. An additional key issue directly relates with the quantification of reliability that predictions of ML models can provide.

In this context, the proposed workshop aims to explore and promote the discussion of techniques fostering explainability and/or reliability estimation of machine learning models.The main objective is to devise potential solutions empowering physicians to effectively apply model predictions in the daily clinical practice.