Anomaly Detection (2)

Using information gleaned from the course concerning a not-well-known distance measure (Gower dissimilarity) useful for cluster models that combine data from mixed variables, including nominal, ordinal, and numeric variables, I was able to refine a model I was working on for an academic conference paper.  More broadly, the subject of identifying anomalies is a central task of academic historians, and this course allowed me the welcome luxury to reflect on their nature and methods of discovery.

Milan Hejtmanek
Seoul National University
Close Menu