« blib/lib/Paws/Comprehend/BatchItemError.pm
↑ Summary
blib/lib/Paws/Comprehend/ClassifierMetadata.pm »
| 1 | # Generated by default/object.tt | ||
| 2 | package Paws::Comprehend::ClassifierEvaluationMetrics; | ||
| 3 | use Moose; | ||
| 4 | has Accuracy => (is => 'ro', isa => 'Num'); | ||
| 5 | has F1Score => (is => 'ro', isa => 'Num'); | ||
| 6 | has HammingLoss => (is => 'ro', isa => 'Num'); | ||
| 7 | has MicroF1Score => (is => 'ro', isa => 'Num'); | ||
| 8 | has MicroPrecision => (is => 'ro', isa => 'Num'); | ||
| 9 | has MicroRecall => (is => 'ro', isa => 'Num'); | ||
| 10 | has Precision => (is => 'ro', isa => 'Num'); | ||
| 11 | has Recall => (is => 'ro', isa => 'Num'); | ||
| 12 | |||
| 13 | 1; | ||
| 14 | |||
| 15 | ### main pod documentation begin ### | ||
| 16 | |||
| 17 | =head1 NAME | ||
| 18 | |||
| 19 | Paws::Comprehend::ClassifierEvaluationMetrics | ||
| 20 | |||
| 21 | =head1 USAGE | ||
| 22 | |||
| 23 | This class represents one of two things: | ||
| 24 | |||
| 25 | =head3 Arguments in a call to a service | ||
| 26 | |||
| 27 | Use the attributes of this class as arguments to methods. You shouldn't make instances of this class. | ||
| 28 | Each attribute should be used as a named argument in the calls that expect this type of object. | ||
| 29 | |||
| 30 | As an example, if Att1 is expected to be a Paws::Comprehend::ClassifierEvaluationMetrics object: | ||
| 31 | |||
| 32 | $service_obj->Method(Att1 => { Accuracy => $value, ..., Recall => $value }); | ||
| 33 | |||
| 34 | =head3 Results returned from an API call | ||
| 35 | |||
| 36 | Use accessors for each attribute. If Att1 is expected to be an Paws::Comprehend::ClassifierEvaluationMetrics object: | ||
| 37 | |||
| 38 | $result = $service_obj->Method(...); | ||
| 39 | $result->Att1->Accuracy | ||
| 40 | |||
| 41 | =head1 DESCRIPTION | ||
| 42 | |||
| 43 | Describes the result metrics for the test data associated with an | ||
| 44 | documentation classifier. | ||
| 45 | |||
| 46 | =head1 ATTRIBUTES | ||
| 47 | |||
| 48 | |||
| 49 | =head2 Accuracy => Num | ||
| 50 | |||
| 51 | The fraction of the labels that were correct recognized. It is computed | ||
| 52 | by dividing the number of labels in the test documents that were | ||
| 53 | correctly recognized by the total number of labels in the test | ||
| 54 | documents. | ||
| 55 | |||
| 56 | |||
| 57 | =head2 F1Score => Num | ||
| 58 | |||
| 59 | A measure of how accurate the classifier results are for the test data. | ||
| 60 | It is derived from the C<Precision> and C<Recall> values. The | ||
| 61 | C<F1Score> is the harmonic average of the two scores. The highest score | ||
| 62 | is 1, and the worst score is 0. | ||
| 63 | |||
| 64 | |||
| 65 | =head2 HammingLoss => Num | ||
| 66 | |||
| 67 | Indicates the fraction of labels that are incorrectly predicted. Also | ||
| 68 | seen as the fraction of wrong labels compared to the total number of | ||
| 69 | labels. Scores closer to zero are better. | ||
| 70 | |||
| 71 | |||
| 72 | =head2 MicroF1Score => Num | ||
| 73 | |||
| 74 | A measure of how accurate the classifier results are for the test data. | ||
| 75 | It is a combination of the C<Micro Precision> and C<Micro Recall> | ||
| 76 | values. The C<Micro F1Score> is the harmonic mean of the two scores. | ||
| 77 | The highest score is 1, and the worst score is 0. | ||
| 78 | |||
| 79 | |||
| 80 | =head2 MicroPrecision => Num | ||
| 81 | |||
| 82 | A measure of the usefulness of the recognizer results in the test data. | ||
| 83 | High precision means that the recognizer returned substantially more | ||
| 84 | relevant results than irrelevant ones. Unlike the Precision metric | ||
| 85 | which comes from averaging the precision of all available labels, this | ||
| 86 | is based on the overall score of all precision scores added together. | ||
| 87 | |||
| 88 | |||
| 89 | =head2 MicroRecall => Num | ||
| 90 | |||
| 91 | A measure of how complete the classifier results are for the test data. | ||
| 92 | High recall means that the classifier returned most of the relevant | ||
| 93 | results. Specifically, this indicates how many of the correct | ||
| 94 | categories in the text that the model can predict. It is a percentage | ||
| 95 | of correct categories in the text that can found. Instead of averaging | ||
| 96 | the recall scores of all labels (as with Recall), micro Recall is based | ||
| 97 | on the overall score of all recall scores added together. | ||
| 98 | |||
| 99 | |||
| 100 | =head2 Precision => Num | ||
| 101 | |||
| 102 | A measure of the usefulness of the classifier results in the test data. | ||
| 103 | High precision means that the classifier returned substantially more | ||
| 104 | relevant results than irrelevant ones. | ||
| 105 | |||
| 106 | |||
| 107 | =head2 Recall => Num | ||
| 108 | |||
| 109 | A measure of how complete the classifier results are for the test data. | ||
| 110 | High recall means that the classifier returned most of the relevant | ||
| 111 | results. | ||
| 112 | |||
| 113 | |||
| 114 | |||
| 115 | =head1 SEE ALSO | ||
| 116 | |||
| 117 | This class forms part of L<Paws>, describing an object used in L<Paws::Comprehend> | ||
| 118 | |||
| 119 | =head1 BUGS and CONTRIBUTIONS | ||
| 120 | |||
| 121 | The source code is located here: L<https://github.com/pplu/aws-sdk-perl> | ||
| 122 | |||
| 123 | Please report bugs to: L<https://github.com/pplu/aws-sdk-perl/issues> | ||
| 124 | |||
| 125 | =cut | ||
| 126 |