binary_log_loss() [2/6]

template<typename T , require_arithmetic_t< T > * = nullptr>

T stan::math::binary_log_loss ( int  y, const T &  y_hat  )

inline

Returns the log loss function for binary classification with specified reference and response values.

The log loss function for prediction $\hat{y} \in [0, 1]$ given outcome $y \in \{ 0, 1 \}$ is

$\mbox{logloss}(1, \hat{y}) = -\log \hat{y}$, and

$\mbox{logloss}(0, \hat{y}) = -\log (1 - \hat{y})$.

Template Parameters

T	value type

Parameters

[in]	y	reference value, either 0 or 1
[in]	y_hat	response value in [0, 1]

Returns

Log loss for response given reference value

Definition at line 30 of file binary_log_loss.hpp.