y_true and y_predy_true values are expected to be -1 or 1. If binary (0 or 1) labels are
provided we will convert them to -1 or 1.
metric_hinge(y_true, y_pred, ..., name = "hinge", dtype = NULL)If y_true and y_pred are missing, a (subclassed) Metric
instance is returned. The Metric object can be passed directly to
compile(metrics = ) or used as a standalone object. See ?Metric for
example usage.
Alternatively, if called with y_true and y_pred arguments, then the
computed case-wise values for the mini-batch are returned directly.
Tensor of true targets.
Tensor of predicted targets.
Passed on to the underlying metric. Used for forwards and backwards compatibility.
(Optional) string name of the metric instance.
(Optional) data type of the metric result.
loss = tf$reduce_mean(tf$maximum(1 - y_true * y_pred, 0L), axis=-1L)
Other metrics:
custom_metric(),
metric_accuracy(),
metric_auc(),
metric_binary_accuracy(),
metric_binary_crossentropy(),
metric_categorical_accuracy(),
metric_categorical_crossentropy(),
metric_categorical_hinge(),
metric_cosine_similarity(),
metric_false_negatives(),
metric_false_positives(),
metric_kullback_leibler_divergence(),
metric_logcosh_error(),
metric_mean_absolute_error(),
metric_mean_absolute_percentage_error(),
metric_mean_iou(),
metric_mean_relative_error(),
metric_mean_squared_error(),
metric_mean_squared_logarithmic_error(),
metric_mean_tensor(),
metric_mean_wrapper(),
metric_mean(),
metric_poisson(),
metric_precision_at_recall(),
metric_precision(),
metric_recall_at_precision(),
metric_recall(),
metric_root_mean_squared_error(),
metric_sensitivity_at_specificity(),
metric_sparse_categorical_accuracy(),
metric_sparse_categorical_crossentropy(),
metric_sparse_top_k_categorical_accuracy(),
metric_specificity_at_sensitivity(),
metric_squared_hinge(),
metric_sum(),
metric_top_k_categorical_accuracy(),
metric_true_negatives(),
metric_true_positives()