{"revision":"c10276e953456835cdbb1f3a60d58ae1980be353","document":{"id":728,"path":"api-reference/server-api-reference/ClassificationSolutionVersionAPI.md","title":"ClassificationSolutionVersion - Global","bundle":null,"doc_type":null,"release":"australia","body":"\n# ClassificationSolutionVersion- Global\n\nThe ClassificationSolutionVersion API is a scriptable object used in Predictive Intelligence stores.\n\nThis API requires the Predictive Intelligence plugin \\(com.glide.platform\\_ml\\) and is provided within the `sn_ml` namespace.\n\nThis API is used for working with solution versions based on [ClassificationSolution API](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/australia/api-reference/server-api-reference/ClassificationSolutionAPI.md) objects in the [ClassificationSolution store](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/australia/api-reference/server-api-reference/ClassificationSolutionStoreAPI.md).\n\nThe system creates a solution version each time you train a solution definition. Most versions are created during scheduled solution training.\n\nMethods in this API are accessible using the following [ClassificationSolution - Global](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/australia/api-reference/server-api-reference/ClassificationSolutionAPI.md) methods:\n\n- [getActiveVersion\\(\\)](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/australia/api-reference/server-api-reference/ClassificationSolutionAPI.md)\n- [getAllVersions\\(\\)](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/australia/api-reference/server-api-reference/ClassificationSolutionAPI.md)\n- [getLatestVersion\\(\\)](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/australia/api-reference/server-api-reference/ClassificationSolutionAPI.md)\n- [getVersion\\(\\)](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/australia/api-reference/server-api-reference/ClassificationSolutionAPI.md)\n\n**Parent Topic:**[Server API reference](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/australia/api-reference/server-api-reference/api-server.md)\n\n## ClassificationSolutionVersion - getPredictionPerformance\\(Object options\\)\n\nCalculates solution precision and coverage.\n\nTo use this method, you must first create settings using the setPredictionSettings\\(\\) method.\n\n\n
\n\nName\n\n\n\nType\n\n\n\nDescription\n\n
\n\noptions\n\n\n\nObject\n\n\n\nSelects a metric to retrieve within a span of dates. For more information, see [Configuring target metrics](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/intelligent-experiences/configuring-target-metrics-trained-classification-solution.md).```\n{\n \"fromDate\": \"String\",\n \"metricName\": \"String\",\n \"toDate\": \"String\"\n}\n```\n\n
\n\noptions.fromDate\n\n\n\nString\n\n\n\nOptional. Start date from which to retrieve metrics in system date time format.\n\n
\n\noptions.metricName\n\n\n\nString\n\n\n\nName of the metric.Valid values:\n\n- **precision**\n- **coverage**\n\n
\n\noptions.toDate\n\n\n\nString\n\n\n\nOptional. End date from which to retrieve metrics.Formnt: System date time format.\n\n
\n
\n\nType\n\n\n\nDescription\n\n
\n\nObject\n\n\n\nParseable JavaScript object containing the results for the input metric.```\n{\n \"metricName\" : \"String\",\n \"metricValue\" : \"String\",\n \"numberOfPredictionsConsidered\" : \"String\"\n}\n```\n\n
\n\nObject.metricName\n\n\n\nName of metric selected as input.Data type: String\n\n
\n\nObject.metricValue\n\n\n\nValue of the selected metric.Data type: String\n\n
\n\nObject.numberOfRecordsConsidered\n\n\n\nNumber of records in the Predictor Results \\[ml\\_predictor\\_results\\] table considered for calculations.Data type: String\n\n
The following example shows how to get a precision value from the active version of a solution.\n\n```\n// Get precision value\nvar mlSolution = sn_ml.ClassificationSolutionStore.get('ml_incident_categorization');\n\ngs.print(JSON.stringify(JSON.parse(mlSolution.getActiveVersion().getPredictionPerformance({\"metricName\": \"precision\"})), null, 2));\n```\n\nOutput:\n\n```\n{\n \"metricName\": \"precision\",\n \"metricValue\": \"70.10\",\n \"numberOfPredictionsConsidered\": \"10\"\n}\n```\n\n## ClassificationSolutionVersion - getPredictionSettings\\(\\)\n\nGets estimated precision, estimated coverage, estimated recall values from the ML Solutions \\[ml\\_solution\\] table and records from the Class Confidence \\[ml\\_class\\] table.\n\n|Name|Type|Description|\n|----|----|-----------|\n|None| | |\n\n\n
\n\nType\n\n\n\nDescription\n\n
\n\nObject\n\n\n\nReturns prediction settings for each class. Data includes precision, coverage, recall, and distribution values of all classes included in the trained model from the ML Solutions \\[ml\\_solution\\] table. For details, see [Predictive model components](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/intelligent-experiences/predictive-intelligence.md).```\n{\n \"\": {Object}\n}\n```\n\n
\n\n<Object>.<class name>\n\n\n\nTraining statistics data from the Class Confidence \\[ml\\_class\\] table. For more information, see [Configuring target metrics](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/intelligent-experiences/configuring-target-metrics-trained-classification-solution.md).Data type: Object\n\n```\n{\n \"precision\": \"String\",\n \"coverage\": \"String\",\n \"recall\": \"String\"\n}\n```\n\n
\n\n<Object>.<class name>.precision\n\n\n\nEstimated solution precision metric for this class.Data type: String \\(numerical value\\)\n\n
\n\n<Object>.<class name>.coverage\n\n\n\nEstimated solution coverage metric for this class.Data type: String \\(numerical value\\)\n\n
\n\n<Object>.<class name>.recall\n\n\n\nEstimated solution recall metric for this class.Data type: String \\(numerical value\\)\n\n
The following example shows how to get prediction settings of an active solution version.\n\n```\nvar mlSolution = sn_ml.ClassificationSolutionStore.get('ml_incident_categorization');\n\ngs.print(JSON.stringify(JSON.parse(mlSolution.getActiveVersion().getPredictionSettings()), null, 2));\n```\n\nOutput:\n\n```\n{\n \"solutionSettings\": {\n \"precision\": \"61.69\",\n \"coverage\": \"66.75\",\n \"recall\": \"57.54\"\n },\n \"classSettings\": {\n \"Approvals\": {\n \"precision\": \"100\",\n \"coverage\": \"0\",\n \"recall\": \"0\"\n },\n \"Authentication\": {\n \"precision\": \"38.71\",\n \"coverage\": \"98.44\",\n \"recall\": \"56.25\"\n },\n \"Automation\": {\n \"precision\": \"100\",\n \"coverage\": \"0\",\n \"recall\": \"0\"\n }, ...\n```\n\n## ClassificationSolutionVersion - getProperties\\(\\)\n\nGets solution object properties and version number.\n\n|Name|Type|Description|\n|----|----|-----------|\n|None| | |\n\n\n
\n\nType\n\n\n\nDescription\n\n
\n\nObject\n\n\n\nContents of the Dataset and [ClassificationSolution](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/australia/api-reference/server-api-reference/ClassificationSolutionAPI.md) version details. Results vary by object property setup.```\n{\n \"algorithmConfig\": {\n \"algorithm\": \"String\",\n \"targetClassRecall\": \"String\"\n },\n \"datasetProperties\": {Object},\n \"domainName\": \"String\",\n \"encoder\": {Object},\n \"groupByFieldName\": \"String\", \n \"inputFieldNames\": [Array],\n \"isActive\": Boolean,\n \"label\": \"String\",\n \"name\": \"String\",\n \"predictedFieldName\": \"String\",\n \"processingLanguage\": \"String\",\n \"scope\": \"String\",\n \"stopwords\": [Array],\n \"stopwords\": [Array],\n \"trainingFrequency\": \"String\",\n \"versionNumber\": \"Number\"\n}\n```\n\n
\n\n<Object>.algorithmConfig\n\n\n\nMethod for encoding the solution. Data type: Object\n\n
\n\n<Object>.algorithmConfig.algorithm\n\n\n\nName of the encoding algorithm for training this solution. Possible values:\n\n- **xgboost**: XGBoost encoding to optimize the training.\n- **logisticRegression**: Method using the logistic regression model for categorical targets such as nominal or ordinal.\n\nData type: String\n\n
\n\n<Object>.algorithmConfig.targetClassRecall\n\n\n\nClass recall parameter to steer a solution's training to bias a specific class. The recall value is a number between 0 and 100 representing a percentage.Data type: String\n\n
\n\n<Object>.datasetProperties\n\n\n\nLists the properties of the [DatasetDefinition\\(\\)](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/australia/api-reference/server-api-reference/DatasetDefinitionAPI.md) object associated with the solution.\n\n Data type: Object\n\n ```\n{\n \"encodedQuery\": \"String\",\n \"fieldDetails\": [Array],\n \"fieldNames\": [Array],\n \"tableName\": \"String\"\n}\n```\n\n
\n\n<Object>.datasetProperties.tableName\n\n\n\nName of the table for the dataset. For example, `\"tableName\" : \"Incident\"`. Data type: String\n\n
\n\n<Object>.datasetProperties.fieldNames\n\n\n\nList of field names from the specified table as strings. For example, `\"fieldNames\" : [\"short_description\", \"priority\"]`. Data type: Array\n\n
\n\n<Object>.datasetProperties.fieldNames.fieldDetails\n\n\n\nList of JavaScript objects that specify field properties. Data type: Array of Objects\n\n```\n[\n {\n \"name\": \"String\",\n \"type\": \"String\"\n }\n]\n```\n\n
\n\n<Object>.datasetProperties.fieldNames.fieldDetails.<object>.name\n\n\n\nName of the field defining the type of information to restrict this dataset to. Data type: String\n\n
\n\n<Object>.datasetProperties.fieldDetails.<object>.type\n\n\n\nMachine-learning field type. Data type: String\n\n
\n\n<Object>.datasetProperties.fieldDetails.encodedQuery\n\n\n\nEncoded query string in the standard platform format. See [Encoded query strings](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/platform-user-interface/c_EncodedQueryStrings.md).Data type: String.\n\n
\n\n<Object>.domainName\n\n\n\nDomain name associated with this dataset. See [Domain separation and Predictive Intelligence](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/intelligent-experiences/domain-separation-predictive-intelligence.md).Type: String\n\n
\n\n<Object>.encoder\n\n\n\nEncoder object assigned to this solution. See [Encoder - Encoder\\(Object config\\)](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/australia/api-reference/server-api-reference/EncoderAPI.md).Data type: Object\n\n
\n\n<Object>.groupByFieldName\n\n\n\nField name by which the system groups records to build classification solutions. Data type: String\n\n
\n\n<Object>.inputFieldNames\n\n\n\nList of input field names as strings. The model uses these fields used to make predictions. Data type: String\n\n
\n\n<Object>.isActive\n\n\n\nFlag that indicates whether this version is active.Valid values:\n\n- true: Version is active.\n- false: Version is not active.\n\n Data type: String\n\n
\n\n<Object>.label\n\n\n\nIdentifies the prediction task. ```\n{\n \"label\": \"my first prediction\"\n}\n```\n\nData type: String\n\n
\n\n<Object>.name\n\n\n\nSystem-assigned name. Data type: String\n\n
\n\n<Object>.predictedFieldName\n\n\n\nIdentifies a field to be trained for predictability. Data type: String\n\n
\n\n<Object>.processingLanguage\n\n\n\nProcessing language in two-letter ISO 639-1 language code format. Data type: String\n\n
\n\n<Object>.scope\n\n\n\nObject scope. Currently the only valid value is `global`.Data type: String\n\n
\n\n<Object>.stopwords\n\n\n\nOptional. Preset list of strings that the system automatically generates based on the **language** property setting. For details, see [Create a custom stopwords list](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/intelligent-experiences/create-custom-stopwords-list.md). Data type: Array\n\n
\n\n<Object>.trainingFrequency\n\n\n\nThe frequency to retrain the model. Possible values:\n\n- every\\_30\\_days\n- every\\_60\\_days\n- every\\_90\\_days\n- every\\_120\\_days\n- every\\_180\\_days\n- run\\_once\n\nDefault: run\\_once\n\nData type: String\n\n
\n\n<Object>.versionNumber\n\n\n\nVersion number of the ClassificationSolution object.Data type: String\n\n
The following example gets properties of the active object version in the store.\n\n```\n// Get properties\nvar mlSolution = sn_ml.ClassificationSolutionStore.get('ml_incident_categorization');\n\ngs.print(JSON.stringify(JSON.parse(mlSolution.getActiveVersion().getProperties()), null, 2));\n```\n\nOutput:\n\n```\n*** Script: {\n \"datasetProperties\": {\n \"encodedQuery\": \"activeANYTHING^EQ\",\n \"fieldNames\": [\n \"short_description\",\n \"category\"\n ],\n \"tableName\": \"incident\"\n },\n \"domainName\": \"global\",\n \"inputFieldNames\": [\n \"short_description\"\n ],\n \"isActive\": \"true\",\n \"label\": \"Incident Categorization_Trainer\",\n \"name\": \"ml_incident_categorization\",\n \"predictedFieldName\": \"category\",\n \"processingLanguage\": \"en\",\n \"stopwords\": [\n \"Default English Stopwords\"\n ],\n \"versionNumber\": \"1\"\n}\n```\n\n## ClassificationSolutionVersion - getStatus\\(Boolean includeDetails\\)\n\nGets training completion status.\n\n\n
\n\nName\n\n\n\nType\n\n\n\nDescription\n\n
\n\nincludeDetails\n\n\n\nBoolean\n\n\n\nFlag that indicates whether to return status **details**.Valid values:\n\n- true: Return additional details.\n- false: Don't return additional details.\n\nDefault: False\n\n
\n
\n\nType\n\n\n\nDescription\n\n
\n\nObject\n\n\n\nJavaScript object containing training status information for a [ClassificationSolution](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/australia/api-reference/server-api-reference/ClassificationSolutionAPI.md) object.```\n{\n \"state\": \"String\",\n \"percentComplete\": \"Number as a String\",\n \"hasJobEnded\": \"Boolean value as a String\",\n \"details\": {Object}\n}\n```\n\n
\n\n<Object>.state\n\n\n\nTraining completion state. If the training job reaches a terminal state, the job does not leave that state. If the state is terminal, the **hasJobEnded** property is set to `true`.Possible values:\n\n- `fetching_files_for_training`\n- `preparing_data`\n- `retry`\n- `solution_cancelled` \\(terminal\\)\n- `solution_complete`\\(terminal\\)\n- `solution_error` \\(terminal\\)\n- `solution_incomplete`\n- `training_request_received`\n- `training_request_timed_out` \\(terminal\\)\n- `training_solution`\n- `uploading_solution`\n- `waiting_for_training`\n\nData type: String\n\n
\n\n<Object>.hasJobEnded\n\n\n\nFlag that indicates whether training is complete.Valid values:\n\n- true: Training is complete.\n- false: Training is incomplete.\n\nData type: Boolean value as a String\n\n
\n\n<Object>.percentComplete\n\n\n\nTraining percent complete. If the completion percentage is less than 100, the job might be in a terminal state. For example, if training times out.Data type: Number as a String\n\nRange: 0 thru 100\n\n
\n\n<Object>.details\n\n\n\nObject containing a list of additional training details.Data type: Object\n\n
The following example shows a successful result with training complete.\n\n```\n// Get status\nvar mlSolution = sn_ml.ClassificationSolutionStore.get('ml_incident_categorization');\n\ngs.print(JSON.stringify(JSON.parse(mlSolution.getActiveVersion().getStatus(true), null, 2)));\n```\n\nOutput:\n\n```\n{\n \"state\":\"solution_complete\",\n \"percentComplete\":\"100\",\n \"hasJobEnded\":\"true\",\n \"details\":{\"stepLabel\":\"Solution Complete\"} // This information is only returned if getStatus(true);\n}\n```\n\nThe following example shows an unsuccessful result with training complete.\n\n```\n// Get status\nvar solutionName = 'ml_x_snc_global_global_classification_solution';\nvar mlSolution = sn_ml.ClassificationSolutionStore.get(solutionName);\nvar trainingStatus = mlSolution.getLatestVersion().getStatus();\n\ngs.print(JSON.stringify(JSON.parse(trainingStatus), null, 2));\n```\n\nOutput:\n\n```\n{\n \"state\":\"solution_error\",\n \"percentComplete\":\"100\",\n \"hasJobEnded\":\"true\"\n}\n```\n\n## ClassificationSolutionVersion - getTrainingStatistics\\(\\)\n\nGets all the training statistics for a given solution.\n\n|Name|Type|Description|\n|----|----|-----------|\n|None| | |\n\n\n
\n\nType\n\n\n\nDescription\n\n
\n\nObject\n\n\n\nJavaScript object with statistics for each class included in training and values for each class that was excluded from training. For details, see [Predictive model components](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/intelligent-experiences/predictive-intelligence.md).```\n{\n \"includedClasses\": {Object},\n \"excludedClasses\": {Object}\n}\n```\n\n
\n\n<Object>.includedClasses\n\n\n\nObject containing precision, coverage, recall, and distribution values of each class in the trained model.Data type: Object\n\n```\n\n \"includedClasses\": {\n \"\": {\n \"distribution\": \"String\",\n \"statistics\": [Array]\n },\n```\n\n
\n\n<Object>.includedClasses.<class name>.distribution\n\n\n\nValue of the outcome probability distribution for this class.Data type: String\n\n
\n\n<Object>.includedClasses.<class name>.statistics\n\n\n\nList of objects providing training statistics data. For more information, see [Configuring target metrics](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/intelligent-experiences/configuring-target-metrics-trained-classification-solution.md).Data type: Array of Objects\n\n```\n[\n {\n \"coverage\": \"String\",\n \"precision\": \"String\",\n \"recall\": \"String\",\n \"selected\": \"String\"\n }\n]\n```\n\n
\n\n<Object>.includedClasses.<class name>.statistics.coverage\n\n\n\nEstimated solution coverage metric for this class.Data type: String \\(numerical value\\)\n\n
\n\n<Object>.includedClasses.<class name>.statistics.precision\n\n\n\nEstimated solution precision metric for this class.Data type: String \\(numerical value\\)\n\n
\n\n<Object>.includedClasses.<class name>.statistics.recall\n\n\n\nEstimated solution recall metric for this class.Data type: String \\(numerical value\\)\n\n
\n\n<Object>.includedClasses.<class name>.statistics.selected\n\n\n\nFlag that indicates whether this class contains records included in solution training.Valid values:\n\n- true: This class contains records selected for training the solution.\n- false: This class is excluded from training.\n\nData type: String \\(Boolean value\\)\n\nDefault: false\n\n
\n\n<Object>.excludedClasses\n\n\n\nDistribution and row count of all the classes excluded from training.Data type: Object\n\n```\n\n \"excludedClasses\": {\n \"\": {\n \"distribution\": \"0.11\",\n \"rowCount\": \"9\"\n }\n```\n\n
\n\n<Object>.excludedClasses.<class name>.distribution\n\n\n\nValue of the outcome probability distribution for this class.Data type: String\n\n
\n\n<Object>.excludedClasses.<class name>.rowCount\n\n\n\nNumber of rows excluded from training.Data type: String \\(numerical value\\)\n\n
The following example shows how to get training statistics from the active solution.\n\n```\n// Get training stats\nvar mlSolution = sn_ml.ClassificationSolutionStore.get('ml_incident_categorization');\n\ngs.print(JSON.stringify(JSON.parse(mlSolution.getActiveVersion().getTrainingStatistics()), null, 2));\n```\n\nOutput:\n\n```\n{\n \"includedClasses\": {\n \"Approvals\": {\n \"distribution\": \"0.43\",\n \"statistics\": [\n {\n \"precision\": \"100\",\n \"coverage\": \"0\",\n \"recall\": \"0\",\n \"selected\": \"true\"\n }\n ]\n }, \n \"Authentication\": {\n \"distribution\": \"4.3\",\n \"statistics\": [\n {\n \"precision\": \"39.33\",\n \"coverage\": \"95.31\",\n \"recall\": \"54.69\",\n \"selected\": \"false\"\n },\n {\n \"precision\": \"100\",\n \"coverage\": \"0\",\n \"recall\": \"0\",\n \"selected\": \"false\"\n },\n ...\n }\n ...\n \"excludedClasses\": {\n \"ACE\": {\n \"distribution\": \"0.11\",\n \"rowCount\": \"9\"\n },\n \"AHA\": {\n \"distribution\": \"0.01\",\n \"rowCount\": \"1\"\n }, \n ...\n}\n```\n\n## ClassificationSolutionVersion - getVersionNumber\\(\\)\n\nGets the version number of a solution object.\n\n|Name|Type|Description|\n|----|----|-----------|\n|None| | |\n\n|Type|Description|\n|----|-----------|\n|String|Version number.|\n\nThe following example shows how to get a version number.\n\n```\n// Get version number\nvar mlSolution = sn_ml.ClassificationSolutionStore.get('ml_incident_categorization');\n\ngs.print(\"Version number: \"+JSON.stringify(JSON.parse(mlSolution.getActiveVersion().getVersionNumber()), null, 2));\n```\n\nOutput:\n\n```\nVersion number: 1\n```\n\n## ClassificationSolutionVersion - predict\\(Object input, Object options\\)\n\nGets the input data for a prediction.\n\n\n
\n\nName\n\n\n\nType\n\n\n\nDescription\n\n
\n\ninput\n\n\n\nObject\n\n\n\n[GlideRecord](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/australia/api-reference/server-api-reference/c_GlideRecordAPI.md) or array of JSON objects containing field names and values as key-value pairs.\n\n
\n\noptions\n\n\n\nObject\n\n\n\nOptional values for filtering prediction results.```\n{\n \"apply_threshold\": Boolean,\n \"top_n\": Number\n}\n```\n\n
\n\noptions.apply\\_threshold\n\n\n\nBoolean\n\n\n\nFlag that indicates whether to check the threshold value for the solution and apply it to the result set.Valid values:\n\n- true: Return results in which confidence is greater than threshold.\n- false: Return all results.\n\nDefault: True\n\n
\n\noptions.top\\_n\n\n\n\nNumber\n\n\n\nIf provided, returns the top results, up to the specified number of predictions.\n\n
\n
\n\nType\n\n\n\nDescription\n\n
\n\nObject\n\n\n\nJSON object containing the prediction results sorted by sys\\_id or record\\_number. ```\n{\n : [Array]\n}\n```\n\n
\n\n<Object>.<identifier>\n\n\n\nList of objects with details for each prediction result.Data type: Array of Objects\n\n```\n: [\n {\n \"confidence\": Number,\n \"predictedSysId\": \"String\",\n \"predictedValue\": \"String\", \n \"threshold\": Number\n }\n]\n```\n\n
\n\n<Object>.<identifier>.<object>.confidence\n\n\n\nValue of the confidence associated with the prediction. For example, 53.84. Data type: Number\n\n
\n\n<Object>.<identifier>.<object>.predictedSysId\n\n\n\nThe sys\\_id of the predicted value. Results can be from any table on which information is being predicted. Data type: String\n\n
\n\n<Object>.<identifier>.<object>.predictedValue\n\n\n\nValue representing the prediction result. Data type: String\n\n
\n\n<Object>.<identifier>.<object>.threshold\n\n\n\nValue of the configured threshold associated with the prediction. Data type: Number\n\n
The following example shows how to display prediction results for a predict\\(\\) method that takes a GlideRecord by sys\\_id for input and includes optional parameters to restrict to top three results and exclude the threshold value.\n\n```\nvar mlSolution = sn_ml.ClassificationSolutionStore.get('ml_incident_categorization');\n\n// single GlideRecord input\nvar input = new GlideRecord(\"incident\");\ninput.get(\"\");\n\n// configure optional parameters\nvar options = {};\noptions.top_n = 3;\noptions.apply_threshold = false;\n\nvar results = mlSolution.getVersion(1).predict(input, options);\n// pretty print JSON results\ngs.print(JSON.stringify(JSON.parse(results), null, 2));\n```\n\nOutput:\n\n```\n{\n \"\": [\n {\n \"confidence\": 62.10782320780268,\n \"threshold\": 20.36,\n \"predictedValue\": \"Clone Issues\",\n \"predictedSysId\": \"\"\n },\n {\n \"confidence\": 6.945237375770391,\n \"threshold\": 16.63,\n \"predictedValue\": \"Instance Administration\",\n \"predictedSysId\": \"\"\n },\n {\n \"confidence\": 5.321061076300759,\n \"threshold\": 23.7,\n \"predictedValue\": \"Administration\",\n \"predictedSysId\": \"\"\n }\n ]\n}\n```\n\nThe following example shows how to display prediction results for a predict\\(\\) method that takes an array of field names as key-value pairs for input and includes optional parameters to restrict to top three results and exclude the threshold value.\n\n```\nvar mlSolution = sn_ml.ClassificationSolutionStore.get('ml_incident_categorization');\n\n// key-value pairs input\nvar input = [{\"short_description\":\"my email is not working\"}, {short_description:\"need help with password\"}];\n\n// configure optional parameters\nvar options = {};\noptions.top_n = 3;\noptions.apply_threshold = false;\nvar results = mlSolution.predict(input, options);\n\n// pretty print JSON results\ngs.print(JSON.stringify(JSON.parse(results), null, 2));\n```\n\nOutput:\n\n```\n{\n \"1\": [\n {\n \"confidence\": 37.5023032262591,\n \"threshold\": 10.72,\n \"predictedValue\": \"Authentication\",\n \"predictedSysId\": \"\"\n },\n {\n \"confidence\": 24.439964862166583,\n \"threshold\": 23.7,\n \"predictedValue\": \"Administration\",\n \"predictedSysId\": \"\"\n },\n {\n \"confidence\": 11.736320486031047,\n \"threshold\": 100,\n \"predictedValue\": \"Security\",\n \"predictedSysId\": \"\"\n }\n ],\n \"2\": [\n {\n \"confidence\": 99,\n \"threshold\": 17.77,\n \"predictedValue\": \"Email\",\n \"predictedSysId\": \"\"\n },\n {\n \"confidence\": 3.182137005157543,\n \"threshold\": 10.72,\n \"predictedValue\": \"Authentication\",\n \"predictedSysId\": \"\"\n },\n {\n \"confidence\": 2.8773826570713514,\n \"threshold\": -1,\n \"predictedValue\": \"Email (I/f)\",\n \"predictedSysId\": \"\"\n }\n ]\n}\n```\n\n## ClassificationSolutionVersion - setPredictionSettings\\(Object options\\)\n\nSets precision, coverage, or recall values at solution level or class level.\n\n\n
\n\nName\n\n\n\nType\n\n\n\nDescription\n\n
\n\noptions\n\n\n\nObject\n\n\n\nMetric configuration values. For information about these features, see [Configuring target metrics](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/intelligent-experiences/configuring-target-metrics-trained-classification-solution.md).```\n{ \t \n \"metricName\" : \"String\",\n \"metricValue\" : \"String\",\n \"className\" : \"String\"\n}\n```\n\n
\n\noptions.metricName\n\n\n\nString\n\n\n\nName of the metric to set.Valid values:\n\n- **coverage**\n- **precision**\n- **recall**\n\n
\n\noptions.metricValue\n\n\n\nString\n\n\n\nNumeric value to assign to the metric.\n\n
\n\noptions.className\n\n\n\nString\n\n\n\nOptional. Name of the class to restrict results to. Use the [getTrainingStatistics\\(\\)](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/australia/api-reference/server-api-reference/ClassificationSolutionVersionAPI.md) method to retrieve a complete list of classes for a solution.\n\n
|Type|Description|\n|----|-----------|\n|None| |\n\nThe following example shows how to set a precision metric to 89.5 for a class named `Clone Issues`.\n\n```\nvar mlSolution = sn_ml.ClassificationSolutionStore.get('ml_x_snc_global_global_classification');\nvar input = {\"metricName\" : \"precision\", \"metricValue\" : \"89.5\", \"className\" : \"Clone Issues\"};\n\nmlSolution.getActiveVersion().setTrainingStatistics(input);\n```\n\n"}}