uprocessing
of complex structures representing knowledge with neural networks
umost NNs are based on vectors, and canšt represent knowledge
easily
urecurrent NNs are more powerful, but also more difficult
to handle
uexperimentation with various types of NNs to evaluate
their suitability
vbio-informatics (drug discovery, genome sequencing)
vknowledge management (ontologies, relationships between documents)
uBackground:
AI, specific domain
uFunding:
None now