Advanced Topics in Case-Based Reasoning

Within CBR there is a growing attention to improved methods for similarity assessment, case adaptation, case learning, and case base maintenance. There is also an increased interest for methods that combine CBR with other reasoning forms, also triggered by the recent BigData focus. Methods that reason from past concrete situations (case-based) is the primary target in the course, but methods that reason from generalized models (model-based) will also be characterized. Integrated reasoning methods that cominbe the two, and address problem solving as well as machine learning targets, will be discussed, and related to developments in the above method areas. The specific set of topics covered will to some extent depend on the interests of the students taking the course.

A. Knowledge:
To get a deeper understanding of how case-specific knowledge can be modelled, learned, and utilized in a combined manner.

B. Skills:
To be able to apply the learned theories and principles for building real world CBR systems for targeted application areas.

C. General Competence:
To get an understanding of how methods within CBR relates to - and can be combined with - other methods, both within in AI and computer science more generally.

Colloquia, diskussions, small projects.