| Previous | Table of Contents | Next |
A MiningFunctionSettings object captures the high level specification input for building a data mining model. The intent of
mining function settings is to allow a user to specify the type of the desired result without having to specify a particular
algorithm.
Although mining function settings allow for the specification of algorithm, if this is omitted, the underlying data mining
system is responsible for selecting the algorithm based on basic user-provided parameters.
Subclasses throw exceptions if invalid algorithm-function pairs are supplied.
Superclasses
ModelElement
Contained Elements
AttributeUsageSet
Attributes
desiredExecutionTimeInMinutes
This attribute indicates the maximum execution time (in minutes) allowed for model building. If NULL, the algorithm determines
for how long the model will build. This is to serve as a hint to the algorithm to adjust model building to meet time constraint.
Vendor implementations map support this to varying degrees, e.g., terminate model build if exceeds this limit, intelligently
adjust algorithm parameters to meet this constaints, or dynamically distribute or parallelize the operation.
type: Integer
multiplicity: exactly one
References
algorithmSettings
The optional algorithm settings attribute provides information on the algorithm to be used as well as specific values for
the parameters used by the algorithm. class: MiningAlgorithmSettings defined by: SettingsRefAlgorithmSettings multiplicity:
zero or one inverse: MiningAlgorithmSettings::settings
attributeUsageSet
This specifies how each attribute as input should be treated by the algorithm. The LogicalAttribute referenced by AttributeUsage
objects must be those in the LogicalData objects used in the same settings.
class: AttributeUsageSetdefined by: SettingsHasAttributeUsageSetmultiplicity: zero or oneinverse: AttributeUsageSet::settings
logicalData
This reference indicates the logical nature of the data to be used for model building with the function. The information provided
by this attribute can be used to validate that the algorithm is correct for the function.
class: LogicalData
defined by: SettingsRefLogicalData
multiplicity: exactly one
inverse: LogicalData::settings