Set-valued Classification with Out-of-distribution Detection for Many Classes

Set-valued classification, a new classification paradigm that aims to identify all the plausible classes that an observation belongs to, improves over the traditional classification paradigms in multiple aspects. Existing set-valued classification methods do not consider the possibility that the test set may contain out-of-distribution data, that is, the emergence of a new class that never appeared in the training data. Moreover, they are computationally expensive when the number of classes is large. We propose a Generalized Prediction Set (GPS) approach …

classification data detection distribution emergence identify multiple observation paradigm possibility set test