classification.proto

// Copyright 2021 Google LLC
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
//     http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.

syntax = "proto3";

package google.cloud.automl.v1;

import "google/api/annotations.proto";

option csharp_namespace = "Google.Cloud.AutoML.V1";
option go_package = "google.golang.org/genproto/googleapis/cloud/automl/v1;automl";
option java_multiple_files = true;
option java_outer_classname = "ClassificationProto";
option java_package = "com.google.cloud.automl.v1";
option php_namespace = "Google\\Cloud\\AutoMl\\V1";
option ruby_package = "Google::Cloud::AutoML::V1";

// Type of the classification problem.
enum ClassificationType {
  // An un-set value of this enum.
  CLASSIFICATION_TYPE_UNSPECIFIED = 0;

  // At most one label is allowed per example.
  MULTICLASS = 1;

  // Multiple labels are allowed for one example.
  MULTILABEL = 2;
}

// Contains annotation details specific to classification.
message ClassificationAnnotation {
  // Output only. A confidence estimate between 0.0 and 1.0. A higher value
  // means greater confidence that the annotation is positive. If a user
  // approves an annotation as negative or positive, the score value remains
  // unchanged. If a user creates an annotation, the score is 0 for negative or
  // 1 for positive.
  float score = 1;
}

// Model evaluation metrics for classification problems.
// Note: For Video Classification this metrics only describe quality of the
// Video Classification predictions of "segment_classification" type.
message ClassificationEvaluationMetrics {
  // Metrics for a single confidence threshold.
  message ConfidenceMetricsEntry {
    // Output only. Metrics are computed with an assumption that the model
    // never returns predictions with score lower than this value.
    float confidence_threshold = 1;

    // Output only. Metrics are computed with an assumption that the model
    // always returns at most this many predictions (ordered by their score,
    // descendingly), but they all still need to meet the confidence_threshold.
    int32 position_threshold = 14;

    // Output only. Recall (True Positive Rate) for the given confidence
    // threshold.
    float recall = 2;

    // Output only. Precision for the given confidence threshold.
    float precision = 3;

    // Output only. False Positive Rate for the given confidence threshold.
    float false_positive_rate = 8;

    // Output only. The harmonic mean of recall and precision.
    float f1_score = 4;

    // Output only. The Recall (True Positive Rate) when only considering the
    // label that has the highest prediction score and not below the confidence
    // threshold for each example.
    float recall_at1 = 5;

    // Output only. The precision when only considering the label that has the
    // highest prediction score and not below the confidence threshold for each
    // example.
    float precision_at1 = 6;

    // Output only. The False Positive Rate when only considering the label that
    // has the highest prediction score and not below the confidence threshold
    // for each example.
    float false_positive_rate_at1 = 9;

    // Output only. The harmonic mean of [recall_at1][google.cloud.automl.v1.ClassificationEvaluationMetrics.ConfidenceMetricsEntry.recall_at1] and [precision_at1][google.cloud.automl.v1.ClassificationEvaluationMetrics.ConfidenceMetricsEntry.precision_at1].
    float f1_score_at1 = 7;

    // Output only. The number of model created labels that match a ground truth
    // label.
    int64 true_positive_count = 10;

    // Output only. The number of model created labels that do not match a
    // ground truth label.
    int64 false_positive_count = 11;

    // Output only. The number of ground truth labels that are not matched
    // by a model created label.
    int64 false_negative_count = 12;

    // Output only. The number of labels that were not created by the model,
    // but if they would, they would not match a ground truth label.
    int64 true_negative_count = 13;
  }

  // Confusion matrix of the model running the classification.
  message ConfusionMatrix {
    // Output only. A row in the confusion matrix.
    message Row {
      // Output only. Value of the specific cell in the confusion matrix.
      // The number of values each row has (i.e. the length of the row) is equal
      // to the length of the `annotation_spec_id` field or, if that one is not
      // populated, length of the [display_name][google.cloud.automl.v1.ClassificationEvaluationMetrics.ConfusionMatrix.display_name] field.
      repeated int32 example_count = 1;
    }

    // Output only. IDs of the annotation specs used in the confusion matrix.
    // For Tables CLASSIFICATION
    // [prediction_type][google.cloud.automl.v1p1beta.TablesModelMetadata.prediction_type]
    // only list of [annotation_spec_display_name-s][] is populated.
    repeated string annotation_spec_id = 1;

    // Output only. Display name of the annotation specs used in the confusion
    // matrix, as they were at the moment of the evaluation. For Tables
    // CLASSIFICATION
    // [prediction_type-s][google.cloud.automl.v1p1beta.TablesModelMetadata.prediction_type],
    // distinct values of the target column at the moment of the model
    // evaluation are populated here.
    repeated string display_name = 3;

    // Output only. Rows in the confusion matrix. The number of rows is equal to
    // the size of `annotation_spec_id`.
    // `row[i].example_count[j]` is the number of examples that have ground
    // truth of the `annotation_spec_id[i]` and are predicted as
    // `annotation_spec_id[j]` by the model being evaluated.
    repeated Row row = 2;
  }

  // Output only. The Area Under Precision-Recall Curve metric. Micro-averaged
  // for the overall evaluation.
  float au_prc = 1;

  // Output only. The Area Under Receiver Operating Characteristic curve metric.
  // Micro-averaged for the overall evaluation.
  float au_roc = 6;

  // Output only. The Log Loss metric.
  float log_loss = 7;

  // Output only. Metrics for each confidence_threshold in
  // 0.00,0.05,0.10,...,0.95,0.96,0.97,0.98,0.99 and
  // position_threshold = INT32_MAX_VALUE.
  // ROC and precision-recall curves, and other aggregated metrics are derived
  // from them. The confidence metrics entries may also be supplied for
  // additional values of position_threshold, but from these no aggregated
  // metrics are computed.
  repeated ConfidenceMetricsEntry confidence_metrics_entry = 3;

  // Output only. Confusion matrix of the evaluation.
  // Only set for MULTICLASS classification problems where number
  // of labels is no more than 10.
  // Only set for model level evaluation, not for evaluation per label.
  ConfusionMatrix confusion_matrix = 4;

  // Output only. The annotation spec ids used for this evaluation.
  repeated string annotation_spec_id = 5;
}