This research area addresses the development of machine learning techniques and systems for reasoning over complex domains involving multiple entities, relationships and constraints. The complexity of the information to be processed and of the tasks to be addressed calls for the design of elaborate architectures and advances in fields such as deep and statistical relational learning, structured prediction, multi-task learning, etc. Prototypical and main application domains are Computer Vision and Multimedia, Natural Language Processing, and Bioinformatics.
The Deep and Structured Machine Learning Research Program gathers the following research groups:
- iKernels Group (https://ikernels-portal.disi.unitn.it/)
- Multimedia and Human Understanding Group (http://mhug.disi.unitn.it/)
- Structured Machine Learning Group (http://sml.disi.unitn.it/)
- KnowDive Group (http://knowdive.disi.unitn.it/)