[CFP] Coming soon - MDPI Mathematics special issue
Sara Narteni
Dear CP Community,
I'm pleased to announce that a new special issue titled "Trustworthy Machine Learning: Explainability, Robustness and Conformal Prediction” is going to be opened soon on MDPI Mathematics. More information on the topics is provided below.
Should you be interested in submitting a contribution, do not hesitate to contact me!
Best regards,
Sara Narteni
Postdoctoral Researcher
National Research Council (CNR-IEIIT), Genoa, Italy
--------
Machine learning (ML) and artificial intelligence (AI) are rapidly spreading across a wide range of domains, including autonomous driving, finance, healthcare, and manufacturing. In safety-critical settings, however, their design and deployment require particular care, as prediction errors may lead to serious consequences.
The Trustworthy AI paradigm has recently emerged to guide the development of ML/AI systems according to principles such as transparency, safety, and robustness.
In this context, growing research efforts focus on explainable AI (XAI) methods to enhance model transparency, as well as uncertainty quantification techniques - such as Conformal Prediction - to provide reliability guarantees to ML models.
This Special Issue aims at pushing research frontiers across Trustworthy ML domain, collecting original contributions within (but not limited to) the following fields:
- New approaches to interpretable and explainable AI
- Evaluation metrics and human-centered assessment of explainability
- Explainability for deep learning and foundation models
- Statistical learning methods for AI reliability
- Distribution-free uncertainty quantification and conformal prediction
- Conformal prediction under distribution shift
- Trade-offs between explainability, robustness, and performance
- Robust and fault-tolerant ML algorithms
- Robustness to data shifts and out-of-distribution detection
- Adversarial robustness