Syntax FACTORY_DEF * ClassifierFactory [FACTORY_NAME ] [INPUT FEATURE_TYPE [ ]* []*]* [GROUP_BY []+] [FLUSH_WHEN_GROUPS_CHANGE (Yes|No)] [INPUT_VALUE ] [METHOD (NATURAL|QUANTILE|EQUAL)] [NUM_CLASSES ] [CLASS_ATTR ] [OUTPUT (CLASSIFIED|) FEATURE_TYPE [ ]* []*]* Overview This factory groups input features into classes based on a numeric attribute value. Class divisions can be determined by a number of different classification schemes. Output features a labeled with an attribute to indicate their class ID. Clauses INPUT_VALUE This parameter is the value to classify the feature by. It must be a finite numerical value for each feature. METHOD NATURAL Uses a 1-dimensional k-means algorithm to determine classes which maximize the difference in values between classes and minimize the difference in values within classes. QUANTILE Divides the input features so that each class has the same number of features. Some classes will have 1 fewer feature than the rest if the number of input features does not evenly divided into the number of classes. EQUAL Divides the input features into classes which have an equal range of values. Some classes may be empty if there are gaps in the input values. NUM_CLASSES This parameter determines the number of desired classes. If the number of classes exceeds the number of features, some classes will be empty. CLASS_ATTR This parameter is the output attribute to populate with the feature's class ID. Output Tags CLASSIFIED Features tagged with their class ID. Features are output in order of their input value attribute. As a consequence, features are also output in class-order. Features with invalid value attributes.