The research aims to develop a single neural-symbolic platform by upgrading the technologies developed up to 3rd year and organically integrating them.
Development of probabilistic programming language and execution tools that perform probabilistic integrated reasoning
A technique that can draw conclusions about unseen knowledge through probabilistic reasoning
A technique that explains the basis and process of drawing conclusions in a format that humans can understand
Technology that enables end-to-end learning by conclusion
Neural-symbolic inference engine interface development that can be used in various field
Development of domain adaption inference technology that enables application of knowledge base extracted from fusion model to specialized domain with little data