gpytorch.metrics

See the metrics tutorial for usage instructions.

metrics.mean_absolute_error(test_y: torch.Tensor)

Mean Absolute Error.

metrics.mean_squared_error(test_y: torch.Tensor, squared: bool = True)

Mean Squared Error.

metrics.mean_standardized_log_loss(test_y: torch.Tensor)

Mean Standardized Log Loss. Reference: Page No. 23, Gaussian Processes for Machine Learning, Carl Edward Rasmussen and Christopher K. I. Williams, The MIT Press, 2006. ISBN 0-262-18253-X

metrics.negative_log_predictive_density(test_y: torch.Tensor)
metrics.quantile_coverage_error(test_y: torch.Tensor, quantile: float = 95.0)

Quantile coverage error.