Recall is a metric of performance for classification algorithms that for binary classification is equal to $ tp / (tp + fn) $ , where $ tp $ and $ fn $ are the numbers of true positives and false negatives respectively. More...

Static Public Member Functions
template < typename MLAlgorithm , typename DataType >
static double	Evaluate (MLAlgorithm &model, const DataType &data, const arma::Row< size_t > &labels)
	Run classification and calculate recall. More...

Static Public Attributes
static const bool	NeedsMinimization = false
	Information for hyper-parameter tuning code. More...

Detailed Description

template<AverageStrategy AS, size_t PositiveClass = 1>
class mlpack::cv::Recall< AS, PositiveClass >

Recall is a metric of performance for classification algorithms that for binary classification is equal to $ tp / (tp + fn) $ , where $ tp $ and $ fn $ are the numbers of true positives and false negatives respectively.

For multiclass classification the recall metric can be used with the following strategies for averaging.

Micro. If there are classes in total, the result is equal to
$(tp_0 + tp_1 + \ldots + tp_N) / (tp_0 + tp_1 + \ldots + tp_N + fn_0 + fn_1 + \ldots + fn_N),$
where and are the numbers of true positives and false negatives respectively for the class (label) .
Macro. If there are classes in total, the result is equal to the mean of the values
$tp_0 / (tp_0 + fn_0), tp_1 / (tp_1 + fn_1), \ldots, tp_N / (tp_N + fn_N),$
where and are the numbers of true positives and false negatives respectively for the class (label) .