Scorecard
Related Doc:
package model
The algorithm name is free-type and can be any description for the specific algorithm that produced the model.
The algorithm name is free-type and can be any description for the specific algorithm that produced the model. This attribute is for information only.
Returns true if there are any missing values of all input fields in the specified series.
Returns true if there are any missing values of all input fields in the specified series.
Common attributes of this model
Common attributes of this model
An informational string describing the technique used by the model designer to establish the baseline scores.
An informational string describing the technique used by the model designer to establish the baseline scores.
A single value to use as the baseline comparison score for all characteristics, when determining reason code ranking.
A single value to use as the baseline comparison score for all characteristics, when determining reason code ranking. Alternatively, unique baseline scores may be set for each individual Characteristic as shown below. This attribute is required only when useReasonCodes is "true" and attribute baselineScore is not given for each Characteristic.
The schema of candidate outputs.
The schema of candidate outputs.
Returns class labels of the specified target.
Returns class labels of the specified target.
The class labels in a classification model.
The class labels in a classification model.
Creates an object of subclass of ModelOutputs that is for writing into an output series.
Returns PMML version as a double value
Returns PMML version as a double value
The data dictionary of this model.
The data dictionary of this model.
Returns all candidates output fields of this model when there is no output specified explicitly.
Encodes the input series.
Encodes the input series.
Returns the field of a given name.
Returns the field of a given name.
FieldNotFoundException if a field with the given name does not exist
Get fields by its usage type: 'active', 'target', 'predicted', 'group' and so on
Get fields by its usage type: 'active', 'target', 'predicted', 'group' and so on
Describe the kind of mining model, e.g., whether it is intended to be used for clustering or for classification.
Describe the kind of mining model, e.g., whether it is intended to be used for clustering or for classification.
Returns the field of a given name, None if a field with the given name does not exist.
The header of this model.
The header of this model.
Implicit referenced derived fields for the sub-model except ones defined in the mining schema.
Implicit referenced derived fields for the sub-model except ones defined in the mining schema.
Returns importances of predictors.
Returns importances of predictors.
The sub-classes can override this method to provide classes of target inside model.
The sub-classes can override this method to provide classes of target inside model.
Initial score contains a value which is added to the overall score whenever partial scores are summed up.
Initial score contains a value which is added to the overall score whenever partial scores are summed up.
Referenced derived fields.
Referenced derived fields.
All input fields in an array.
All input fields in an array.
All input names in an array.
All input names in an array.
The schema of inputs.
The schema of inputs.
Tests if this is a association rules model.
Tests if this is a association rules model.
Tests if the target is a binary field
Tests if the target is a binary field
Tests if this is a classification model of the specified target, it's applicable for multiple targets.
Tests if this is a classification model of the specified target, it's applicable for multiple targets.
Tests if this is a classification model.
Tests if this is a classification model.
Tests if this is a clustering model.
Tests if this is a clustering model.
Tests if this is a mixed model.
Tests if this is a mixed model.
Tests if the target is an ordinal field
Tests if the target is an ordinal field
Tests if this is a regression model of the specified target, it's applicable for multiple targets.
Tests if this is a regression model of the specified target, it's applicable for multiple targets.
Tests if this is a regression model.
Tests if this is a regression model.
Indicates if the model is valid for scoring.
Indicates if the model is valid for scoring. If this attribute is true or if it is missing, then the model should be processed normally. However, if the attribute is false, then the model producer has indicated that this model is intended for information purposes only and should not be used to generate results.
Tests if this is a sequences model.
Tests if this is a sequences model.
Tests if this is a time series model.
Tests if this is a time series model.
The optional local transformations.
The optional local transformations.
Model element type.
Identifies the model with a unique name in the context of the PMML file.
Identifies the model with a unique name in the context of the PMML file. This attribute is not required. Consumers of PMML models are free to manage the names of the models at their discretion.
A series with all null values is returned when can not produce a result.
A series with all null values is returned when can not produce a result.
Returns the number of class labels of the specified target.
Returns the number of class labels of the specified target.
The number of class labels in a classification model.
The number of class labels in a classification model.
Returns optype of the specified target.
Returns optype of the specified target.
When Target specifies optype then it overrides the optype attribute in a corresponding MiningField, if it exists.
When Target specifies optype then it overrides the optype attribute in a corresponding MiningField, if it exists. If the target does not specify optype then the MiningField is used as default. And, in turn, if the MiningField does not specify an optype, it is taken from the corresponding DataField. In other words, a MiningField overrides a DataField, and a Target overrides a MiningField.
The schema of final outputs.
The schema of final outputs.
The parent model.
The scoring procedure for a scorecard is simple.
The scoring procedure for a scorecard is simple. Partial scores are summed up to create an overall score, the result of the scorecard.
Predicts one or multiple records in json format, there are two formats supported:
Predicts one or multiple records in json format, there are two formats supported:
- ‘records’ : list like [{column -> value}, … , {column -> value}] - ‘split’ : dict like {‘columns’ -> [columns], ‘data’ -> [values]}
Records in json
Results in json
Predicts values for a given Array, and the order of those values is supposed as same as the input fields list
Predicts values for a given Array, and the order of those values is supposed as same as the input fields list
Predicts values for a given list of key/value pairs.
Predicts values for a given list of key/value pairs.
Predicts values for a given data map of Java.
Predicts values for a given data map of Java.
Predicts values for a given data map.
Predicts values for a given data map.
Pre-process the input series.
Pre-process the input series.
Tests if probabilities of categories of target can be produced by this model.
Tests if probabilities of categories of target can be produced by this model.
May be "pointsAbove" or "pointsBelow", describing how reason codes shall be ranked, relative to the baseline score of each Characteristic, or as set at the top-level scorecard.
May be "pointsAbove" or "pointsBelow", describing how reason codes shall be ranked, relative to the baseline score of each Characteristic, or as set at the top-level scorecard.
Collected all reason codes with order that appears in the PMML file, from top to bottom.
The number of reason codes need to return.
The class labels of all categorical targets.
The class labels of all categorical targets.
The first target field for the supervised model.
The first target field for the supervised model.
All target fields in an array.
All target fields in an array. Multiple target fields are allowed. It depends on the kind of the model whether prediction of multiple fields is supported.
Returns targets that are residual values to be computed, the input data must include target values.
Returns targets that are residual values to be computed, the input data must include target values.
Name of the first target for the supervised model.
Name of the first target for the supervised model.
All target names in an array.
All target names in an array.
The optional transformation dictionary.
The optional transformation dictionary.
If "false", reason codes are not computed as part of the scorecard.
If "false", reason codes are not computed as part of the scorecard.
Setup indices to retrieve data from series faster by index instead of name, the index is immutable when model is built because the model object could run in multiple threads, so it's important make sure the model object is totally immutable.
Setup indices to retrieve data from series faster by index instead of name, the index is immutable when model is built because the model object could run in multiple threads, so it's important make sure the model object is totally immutable.
Setup indices of targets that are usually not used by the scoring process, they are only used when residual values to be computed.
PMML version.
PMML version.
A data mining model contains a set of input fields which are used to predict a certain target value. This prediction can be seen as an assessment about a prospect, a customer, or a scenario for which an outcome is predicted based on historical data. In a scorecard, input fields, also referred to as characteristics (for example, "age"), are broken down into attributes (for example, "19-29" and "30-39" age groups or ranges) with specific partial scores associated with them. These scores represent the influence of the input attributes on the target and are readily available for inspection. Partial scores are then summed up so that an overall score can be obtained for the target value.
Scorecards are very popular in the financial industry for their interpretability and ease of implementation, and because input attributes can be mapped to a series of reason codes which provide explanations of each individual's score. Usually, the lower the overall score produced by a scorecard, the higher the chances of it triggering an adverse decision, which usually involves the referral or denial of services. Reason codes, as the name suggests, allow for an explanation of scorecard behavior and any adverse decisions generated as a consequence of the overall score. They basically answer the question: "Why is the score low, given its input conditions?"