Module trustML.metrics.metric
Expand source code
class Metric:
"""
Metric abstract class
"""
def __init__(self):
self.score = None
def assess(self, trained_model, data_x, data_y):
"""Assessment of the metric using the trained model, dataset predictors and targets passed as parameters.
"""
passClasses
class Metric-
Metric abstract class
Expand source code
class Metric: """ Metric abstract class """ def __init__(self): self.score = None def assess(self, trained_model, data_x, data_y): """Assessment of the metric using the trained model, dataset predictors and targets passed as parameters. """ passSubclasses
- AccuracySKL
- AverageBoundSKLTree
- InvertedBrierSKL
- EqualOpportunitySKL
- InvertedExpectedCalibrationSKL
- ExplanationsAccuracyTED
- F1SKL
- FaithfulnessLIMESKL
- MonotonicityLIMESKL
- PPercentageSKL
- PrecisionSKL
- RecallSKL
- ROCSKL
- VerifiedErrorSKLTree
Methods
def assess(self, trained_model, data_x, data_y)-
Assessment of the metric using the trained model, dataset predictors and targets passed as parameters.
Expand source code
def assess(self, trained_model, data_x, data_y): """Assessment of the metric using the trained model, dataset predictors and targets passed as parameters. """ pass