CLogLoss implements the logarithmic loss function.
Definition at line 21 of file LogLoss.h.
Public Member Functions | |
CLogLoss () | |
~CLogLoss () | |
float64_t | loss (float64_t prediction, float64_t label) |
virtual float64_t | first_derivative (float64_t prediction, float64_t label) |
virtual float64_t | second_derivative (float64_t prediction, float64_t label) |
virtual float64_t | get_update (float64_t prediction, float64_t label, float64_t eta_t, float64_t norm) |
virtual float64_t | get_square_grad (float64_t prediction, float64_t label) |
virtual ELossType | get_loss_type () |
virtual const char * | get_name () const |
Get first derivative of the loss function
prediction | prediction | |
label | label |
Implements CLossFunction.
Definition at line 23 of file LogLoss.cpp.
virtual ELossType get_loss_type | ( | ) | [virtual] |
virtual const char* get_name | ( | ) | const [virtual] |
Get square of gradient, used for adaptive learning
prediction | prediction | |
label | label |
Implements CLossFunction.
Definition at line 66 of file LogLoss.cpp.
float64_t get_update | ( | float64_t | prediction, | |
float64_t | label, | |||
float64_t | eta_t, | |||
float64_t | norm | |||
) | [virtual] |
Get importance aware weight update for this loss function
prediction | prediction | |
label | label | |
eta_t | learning rate at update number t | |
norm | scale value |
Implements CLossFunction.
Definition at line 40 of file LogLoss.cpp.
Get loss for an example
prediction | prediction | |
label | label |
Implements CLossFunction.
Definition at line 15 of file LogLoss.cpp.
Get second derivative of the loss function
prediction | prediction | |
label | label |
Implements CLossFunction.
Definition at line 32 of file LogLoss.cpp.