The AFT loss function.
More...
#include <survival_util.h>
|
| AFTLoss (ProbabilityDistributionType dist_type) |
| Constructor for AFT loss function. More...
|
|
double | Loss (double y_lower, double y_upper, double y_pred, double sigma) |
| Compute the AFT loss. More...
|
|
double | Gradient (double y_lower, double y_upper, double y_pred, double sigma) |
| Compute the gradient of the AFT loss. More...
|
|
double | Hessian (double y_lower, double y_upper, double y_pred, double sigma) |
| Compute the hessian of the AFT loss. More...
|
|
◆ AFTLoss()
Constructor for AFT loss function.
- Parameters
-
dist_type | Choice of probability distribution for the noise term in AFT |
◆ Gradient()
double xgboost::common::AFTLoss::Gradient |
( |
double |
y_lower, |
|
|
double |
y_upper, |
|
|
double |
y_pred, |
|
|
double |
sigma |
|
) |
| |
Compute the gradient of the AFT loss.
- Parameters
-
y_lower | Lower bound for the true label |
y_upper | Upper bound for the true label |
y_pred | Predicted label |
sigma | Scaling factor to be applied to the distribution of the noise term |
◆ Hessian()
double xgboost::common::AFTLoss::Hessian |
( |
double |
y_lower, |
|
|
double |
y_upper, |
|
|
double |
y_pred, |
|
|
double |
sigma |
|
) |
| |
Compute the hessian of the AFT loss.
- Parameters
-
y_lower | Lower bound for the true label |
y_upper | Upper bound for the true label |
y_pred | Predicted label |
sigma | Scaling factor to be applied to the distribution of the noise term |
◆ Loss()
double xgboost::common::AFTLoss::Loss |
( |
double |
y_lower, |
|
|
double |
y_upper, |
|
|
double |
y_pred, |
|
|
double |
sigma |
|
) |
| |
Compute the AFT loss.
- Parameters
-
y_lower | Lower bound for the true label |
y_upper | Upper bound for the true label |
y_pred | Predicted label |
sigma | Scaling factor to be applied to the distribution of the noise term |
The documentation for this class was generated from the following file: