Asymmetric Loss Functions for Forecasting in Criminal Justice Settings

53 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Interpretable classification models for recidivism prediction (2015, methodology): “In this paper, we use a cost-sensitive approach to produce classiﬁcation models at different points of the ROC curve (see, e.g., Berk 2010, 2011).”
Balancing the Costs of Forecasting Errors in Parole Decisions (2010, background): “If the costs are not really equal, the forecasts can be misleading, often badly so (Berk, 2010).”
Forecasts of Violence to Inform Sentencing Decisions (2013, background): “A key asset of classiﬁcation trees is the capacity to take into account the relative costs of false positives and false negatives in the classes assigned (Breiman et al. 1984; Berk 2011).”
A Signaling Perspective on Employment-Based Reentry Programming Training Completion as a Desistance Signal (2012, unknown): “The relative costs of type I and type II errors, which are context specific, drive whether a prediction exercise is justifiable from a policy perspective (Berk, 2011; Bushway, 2011).”
Statistical Procedures for Forecasting Criminal Behavior (2013, background): “Forecasting errors that do not have equal costs can be introduced into the procedure at the beginning so that all of the results properly represent the preferences of stakeholders (Berk, 2011).”