github-mirrors/elasticsearch
1
0
Fork 0
mirror of https://github.com/elastic/elasticsearch.git synced 2025-04-24 23:27:25 -04:00
Code Issues Projects Releases Packages Wiki Activity

Add MlClientDocumentationIT tests for classification. (#47569)

Browse source
This commit is contained in:
Przemysław Witek 2019-10-11 08:21:45 +02:00 • committed by GitHub
parent 95a3da21da
commit 9b5770da0e
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
4 changed files with 272 additions and 40 deletions
Download patch file Download diff file Expand all files Collapse all files

95
docs/java-rest/high-level/ml/evaluate-data-frame.asciidoc
View file

@ -20,14 +20,52 @@ include-tagged::{doc-tests-file}[{api}-request]
<1> Constructing a new evaluation request
<2> Reference to an existing index
<3> The query with which to select data from indices
<4> Kind of evaluation to perform
<5> Name of the field in the index. Its value denotes the actual (i.e. ground truth) label for an example. Must be either true or false
<6> Name of the field in the index. Its value denotes the probability (as per some ML algorithm) of the example being classified as positive
<7> The remaining parameters are the metrics to be calculated based on the two fields described above.
<8> https://en.wikipedia.org/wiki/Precision_and_recall[Precision] calculated at thresholds: 0.4, 0.5 and 0.6
<9> https://en.wikipedia.org/wiki/Precision_and_recall[Recall] calculated at thresholds: 0.5 and 0.7
<10> https://en.wikipedia.org/wiki/Confusion_matrix[Confusion matrix] calculated at threshold 0.5
<11> https://en.wikipedia.org/wiki/Receiver_operating_characteristic#Area_under_the_curve[AuC ROC] calculated and the curve points returned
<4> Evaluation to be performed
==== Evaluation
Evaluation to be performed.
Currently, supported evaluations include: +BinarySoftClassification+, +Classification+, +Regression+.
===== Binary soft classification
["source","java",subs="attributes,callouts,macros"]
--------------------------------------------------
include-tagged::{doc-tests-file}[{api}-evaluation-softclassification]
--------------------------------------------------
<1> Constructing a new evaluation
<2> Name of the field in the index. Its value denotes the actual (i.e. ground truth) label for an example. Must be either true or false.
<3> Name of the field in the index. Its value denotes the probability (as per some ML algorithm) of the example being classified as positive.
<4> The remaining parameters are the metrics to be calculated based on the two fields described above
<5> https://en.wikipedia.org/wiki/Precision_and_recall#Precision[Precision] calculated at thresholds: 0.4, 0.5 and 0.6
<6> https://en.wikipedia.org/wiki/Precision_and_recall#Recall[Recall] calculated at thresholds: 0.5 and 0.7
<7> https://en.wikipedia.org/wiki/Confusion_matrix[Confusion matrix] calculated at threshold 0.5
<8> https://en.wikipedia.org/wiki/Receiver_operating_characteristic#Area_under_the_curve[AuC ROC] calculated and the curve points returned
===== Classification
["source","java",subs="attributes,callouts,macros"]
--------------------------------------------------
include-tagged::{doc-tests-file}[{api}-evaluation-classification]
--------------------------------------------------
<1> Constructing a new evaluation
<2> Name of the field in the index. Its value denotes the actual (i.e. ground truth) class the example belongs to.
<3> Name of the field in the index. Its value denotes the predicted (as per some ML algorithm) class of the example.
<4> The remaining parameters are the metrics to be calculated based on the two fields described above
<5> Multiclass confusion matrix of size 3
===== Regression
["source","java",subs="attributes,callouts,macros"]
--------------------------------------------------
include-tagged::{doc-tests-file}[{api}-evaluation-regression]
--------------------------------------------------
<1> Constructing a new evaluation
<2> Name of the field in the index. Its value denotes the actual (i.e. ground truth) value for an example.
<3> Name of the field in the index. Its value denotes the predicted (as per some ML algorithm) value for the example.
<4> The remaining parameters are the metrics to be calculated based on the two fields described above
<5> https://en.wikipedia.org/wiki/Mean_squared_error[Mean squared error]
<6> https://en.wikipedia.org/wiki/Coefficient_of_determination[R squared]
include::../execution.asciidoc[]
@ -41,7 +79,40 @@ The returned +{response}+ contains the requested evaluation metrics.
include-tagged::{doc-tests-file}[{api}-response]
--------------------------------------------------
<1> Fetching all the calculated metrics results
<2> Fetching precision metric by name
<3> Fetching precision at a given (0.4) threshold
<4> Fetching confusion matrix metric by name
<5> Fetching confusion matrix at a given (0.5) threshold
==== Results
===== Binary soft classification
["source","java",subs="attributes,callouts,macros"]
--------------------------------------------------
include-tagged::{doc-tests-file}[{api}-results-softclassification]
--------------------------------------------------
<1> Fetching precision metric by name
<2> Fetching precision at a given (0.4) threshold
<3> Fetching confusion matrix metric by name
<4> Fetching confusion matrix at a given (0.5) threshold
===== Classification
["source","java",subs="attributes,callouts,macros"]
--------------------------------------------------
include-tagged::{doc-tests-file}[{api}-results-classification]
--------------------------------------------------
<1> Fetching multiclass confusion matrix metric by name
<2> Fetching the contents of the confusion matrix
<3> Fetching the number of classes that were not included in the matrix
===== Regression
["source","java",subs="attributes,callouts,macros"]
--------------------------------------------------
include-tagged::{doc-tests-file}[{api}-results-regression]
--------------------------------------------------
<1> Fetching mean squared error metric by name
<2> Fetching the actual mean squared error value
<3> Fetching R squared metric by name
<4> Fetching the actual R squared value

20
docs/java-rest/high-level/ml/put-data-frame-analytics.asciidoc
View file

@ -76,7 +76,7 @@ include-tagged::{doc-tests-file}[{api}-dest-config]
==== Analysis
The analysis to be performed.
Currently, the supported analyses include : +OutlierDetection+, +Regression+.
Currently, the supported analyses include: +OutlierDetection+, +Classification+, +Regression+.
===== Outlier detection
@ -101,6 +101,24 @@ include-tagged::{doc-tests-file}[{api}-outlier-detection-customized]
<6> The proportion of the data set that is assumed to be outlying prior to outlier detection
<7> Whether to apply standardization to feature values
===== Classification
+Classification+ analysis requires to set which is the +dependent_variable+ and
has a number of other optional parameters:
["source","java",subs="attributes,callouts,macros"]
--------------------------------------------------
include-tagged::{doc-tests-file}[{api}-classification]
--------------------------------------------------
<1> Constructing a new Classification builder object with the required dependent variable
<2> The lambda regularization parameter. A non-negative double.
<3> The gamma regularization parameter. A non-negative double.
<4> The applied shrinkage. A double in [0.001, 1].
<5> The maximum number of trees the forest is allowed to contain. An integer in [1, 2000].
<6> The fraction of features which will be used when selecting a random bag for each candidate split. A double in (0, 1].
<7> The name of the prediction field in the results object.
<8> The percentage of training-eligible rows to be used in training. Defaults to 100%.
===== Regression
+Regression+ analysis requires to set which is the +dependent_variable+ and