View source on GitHub |
Ops for evaluation metrics and summary statistics.
See the Contrib Metrics guide.
Functions
accuracy(...)
: Computes the percentage of times that predictions matches labels.
aggregate_metric_map(...)
: Aggregates the metric names to tuple dictionary.
aggregate_metrics(...)
: Aggregates the metric value tensors and update ops into two lists.
auc_using_histogram(...)
: AUC computed by maintaining histograms.
auc_with_confidence_intervals(...)
: Computes the AUC and asymptotic normally distributed confidence interval.
cohen_kappa(...)
: Calculates Cohen's kappa.
confusion_matrix(...)
: Deprecated. Use tf.math.confusion_matrix instead.
count(...)
: Computes the number of examples, or sum of weights
.
f1_score(...)
: Computes the approximately best F1-score across different thresholds.
precision_at_recall(...)
: Computes the precision at a given recall.
precision_recall_at_equal_thresholds(...)
: A helper method for creating metrics related to precision-recall curves.
recall_at_precision(...)
: Computes recall
at precision
.
set_difference(...)
: Compute set difference of elements in last dimension of a
and b
.
set_intersection(...)
: Compute set intersection of elements in last dimension of a
and b
.
set_size(...)
: Compute number of unique elements along last dimension of a
.
set_union(...)
: Compute set union of elements in last dimension of a
and b
.
sparse_recall_at_top_k(...)
: Computes recall@k of top-k predictions with respect to sparse labels.
streaming_accuracy(...)
: Calculates how often predictions
matches labels
. (deprecated)
streaming_auc(...)
: Computes the approximate AUC via a Riemann sum. (deprecated)
streaming_concat(...)
: Concatenate values along an axis across batches.
streaming_covariance(...)
: Computes the unbiased sample covariance between predictions
and labels
.
streaming_curve_points(...)
: Computes curve (ROC or PR) values for a prespecified number of points.
streaming_dynamic_auc(...)
: Computes the apporixmate AUC by a Riemann sum with data-derived thresholds.
streaming_false_negative_rate(...)
: Computes the false negative rate of predictions with respect to labels.
streaming_false_negative_rate_at_thresholds(...)
: Computes various fnr values for different thresholds
on predictions
.
streaming_false_negatives(...)
: Computes the total number of false negatives. (deprecated)
streaming_false_negatives_at_thresholds(...)
streaming_false_positive_rate(...)
: Computes the false positive rate of predictions with respect to labels.
streaming_false_positive_rate_at_thresholds(...)
: Computes various fpr values for different thresholds
on predictions
.
streaming_false_positives(...)
: Sum the weights of false positives. (deprecated)
streaming_false_positives_at_thresholds(...)
streaming_mean(...)
: Computes the (weighted) mean of the given values. (deprecated)
streaming_mean_absolute_error(...)
: Computes the mean absolute error between the labels and predictions. (deprecated)
streaming_mean_cosine_distance(...)
: Computes the cosine distance between the labels and predictions.
streaming_mean_iou(...)
: Calculate per-step mean Intersection-Over-Union (mIOU).
streaming_mean_relative_error(...)
: Computes the mean relative error by normalizing with the given values.
streaming_mean_squared_error(...)
: Computes the mean squared error between the labels and predictions. (deprecated)
streaming_mean_tensor(...)
: Computes the element-wise (weighted) mean of the given tensors. (deprecated)
streaming_pearson_correlation(...)
: Computes Pearson correlation coefficient between predictions
, labels
.
streaming_percentage_less(...)
: Computes the percentage of values less than the given threshold.
streaming_precision(...)
: Computes the precision of the predictions with respect to the labels. (deprecated)
streaming_precision_at_thresholds(...)
: Computes precision values for different thresholds
on predictions
. (deprecated)
streaming_recall(...)
: Computes the recall of the predictions with respect to the labels. (deprecated)
streaming_recall_at_k(...)
: Computes the recall@k of the predictions with respect to dense labels. (deprecated)
streaming_recall_at_thresholds(...)
: Computes various recall values for different thresholds
on predictions
. (deprecated)
streaming_root_mean_squared_error(...)
: Computes the root mean squared error between the labels and predictions. (deprecated)
streaming_sensitivity_at_specificity(...)
: Computes the sensitivity at a given specificity.
streaming_sparse_average_precision_at_k(...)
: Computes average precision@k of predictions with respect to sparse labels.
streaming_sparse_average_precision_at_top_k(...)
: Computes average precision@k of predictions with respect to sparse labels.
streaming_sparse_precision_at_k(...)
: Computes precision@k of the predictions with respect to sparse labels.
streaming_sparse_precision_at_top_k(...)
: Computes precision@k of top-k predictions with respect to sparse labels.
streaming_sparse_recall_at_k(...)
: Computes recall@k of the predictions with respect to sparse labels.
streaming_specificity_at_sensitivity(...)
: Computes the specificity at a given sensitivity.
streaming_true_negatives(...)
: Sum the weights of true_negatives. (deprecated)
streaming_true_negatives_at_thresholds(...)
streaming_true_positives(...)
: Sum the weights of true_positives. (deprecated)