go back
go back
Volume 14, No. 12
How Divergent Is Your Data?
Authors:
Eliana Pastor (Politecnico di Torino), Andrew Gavgavian (University of California, Santa Cruz), Elena Baralis (Dipartimento di Automatica e Informatica Politecnico di Torino), Luca de Alfaro (University of California, Santa Cruz)
Abstract
We present DivExplorer, a tool that enables users to explore datasets and find subgroups of data for which a classifier behaves in an anomalous manner. These subgroups, denoted as divergent subgroups, may exhibit, for example, higher-than-normal false positive or negative rates. DivExplorer can be used to analyze and debug classifiers. If the data has ethical or social implications, DivExplorer can be also used to identify bias in classifiers.
PVLDB is part of the VLDB Endowment Inc.
Privacy Policy