Skip to yearly menu bar Skip to main content


Poster

Multiclass Performance Metric Elicitation

Gaurush Hiranandani · Shant Boodaghians · Ruta Mehta · Sanmi Koyejo

East Exhibition Hall B, C #226

Keywords: [ Learning Theory ] [ Theory ] [ Algorithms -> Active Learning; Algorithms -> Classification; Algorithms ] [ Ranking and Preference Learning ]


Abstract:

Metric Elicitation is a principled framework for selecting the performance metric that best reflects implicit user preferences. However, available strategies have so far been limited to binary classification. In this paper, we propose novel strategies for eliciting multiclass classification performance metrics using only relative preference feedback. We also show that the strategies are robust to both finite sample and feedback noise.

Live content is unavailable. Log in and register to view live content