applicable models

probabilistic output

Model must provide probability distributions as output

10 techniques
GoalsModelsData TypesDescription
Monte Carlo Dropout
Algorithmic
Architecture/neural Networks
Paradigm/probabilistic
+4
Any
Monte Carlo Dropout estimates prediction uncertainty by applying dropout (randomly setting neural network weights to...
Out-of-Distribution Detector for Neural Networks
Algorithmic
Architecture/neural Networks
Paradigm/discriminative
+3
Any
ODIN (Out-of-Distribution Detector for Neural Networks) identifies when a neural network encounters inputs significantly...
Empirical Calibration
Algorithmic
Architecture/model Agnostic
Paradigm/supervised
+2
Any
Empirical calibration adjusts a model's predicted probabilities to match observed frequencies. For example, if events...
Temperature Scaling
Algorithmic
Architecture/neural Networks
Paradigm/discriminative
+3
Any
Temperature scaling adjusts a model's confidence by applying a single parameter (temperature) to its predictions. When a...
Area Under Precision-Recall Curve
Algorithmic
Architecture/model Agnostic
Paradigm/supervised
+2
Any
Area Under Precision-Recall Curve (AUPRC) measures model performance by plotting precision (the proportion of positive...
Confidence Thresholding
Algorithmic
Architecture/model Agnostic
Requirements/black Box
+1
Any
Confidence thresholding creates decision boundaries based on model uncertainty scores, routing predictions into...
Equalised Odds Post-Processing
Algorithmic
Architecture/model Agnostic
Paradigm/discriminative
+3
Any
A post-processing fairness technique based on Hardt et al.'s seminal work that adjusts classification thresholds after...
Threshold Optimiser
Algorithmic
Architecture/model Agnostic
Paradigm/discriminative
+3
Any
Threshold Optimiser adjusts decision thresholds for different demographic groups after model training to satisfy...
Reject Option Classification
Algorithmic
Architecture/model Agnostic
Paradigm/supervised
+2
Any
A post-processing fairness technique that modifies predictions in regions of high uncertainty to favour disadvantaged...
Calibration with Equality of Opportunity
Algorithmic
Architecture/model Agnostic
Paradigm/supervised
+2
Any
A post-processing fairness technique that adjusts model predictions to achieve equal true positive rates across...
Rows per page
Page 1 of 1