Small jpmml-evaulator demo (with a dummy model) on how to dynamically select a regression model based on an input attribute value. In this case we use the "attribute1" to select the right model.

dynamic-model-selection-example.pmml.xml

<?xml version="1.0" encoding="UTF-8"?>
<PMML version="4.2"
      xmlns="http://www.dmg.org/PMML-4_2"
        >
    <Header copyright="Copyright (c) 2014 tom" description="Linear Regression Model">
        <Extension name="user" value="tom" extender="Rattle/PMML"/>
        <Application name="Rattle/PMML" version="1.3"/>
        <Timestamp>2014-03-15 13:18:06</Timestamp>
    </Header>

    <DataDictionary numberOfFields="3">
        <DataField name="rate" optype="continuous" dataType="double"/>
        <DataField name="year" optype="continuous" dataType="double"/>
        <DataField name="attribute1" optype="categorical" dataType="string"/>
    </DataDictionary>

    <MiningModel functionName="regression">

        <MiningSchema>
            <MiningField name="rate" usageType="predicted"/>
            <MiningField name="year" usageType="active"/>
            <MiningField name="attribute1" usageType="active"/>
        </MiningSchema>

        <Segmentation multipleModelMethod="selectFirst">

            <Segment id="A">
                <SimplePredicate field="attribute1" operator="equal" value="1315"/>

                <RegressionModel modelName="interest-rate-simple-linear-regression-1" functionName="regression"
                                 algorithmName="least squares"
                                 targetFieldName="rate">

                    <MiningSchema>
                        <MiningField name="rate" usageType="predicted"/>
                        <MiningField name="year" usageType="active"/>
                        <MiningField name="attribute1" usageType="active"/>
                    </MiningSchema>

                    <Output>
                        <OutputField name="Predicted_rate" feature="predictedValue"/>
                    </Output>
                    <RegressionTable intercept="10000">
                        <NumericPredictor name="year" exponent="1" coefficient="1.5"/>
                    </RegressionTable>
                </RegressionModel>
            </Segment>

            <Segment id="B">
                <SimplePredicate field="attribute1" operator="equal" value="1330"/>

                <RegressionModel modelName="interest-rate-simple-linear-regression-2" functionName="regression"
                                 algorithmName="least squares"
                                 targetFieldName="rate">

                 <MiningSchema>
                     <MiningField name="rate" usageType="predicted"/>
                     <MiningField name="year" usageType="active"/>
                     <MiningField name="attribute1" usageType="active"/>
                 </MiningSchema>

                 <Output>
                     <OutputField name="Predicted_rate" feature="predictedValue"/>
                 </Output>
                 <RegressionTable intercept="20000">
                     <NumericPredictor name="year" exponent="1" coefficient="-0.5"/>
                 </RegressionTable>
             </RegressionModel>
         </Segment>

         <Segment id="C">

             <!--default model -->

                <True/>

                <RegressionModel modelName="interest-rate-simple-linear-regression-2" functionName="regression"
                                 algorithmName="least squares"
                                 targetFieldName="rate">


                    <MiningSchema>
                        <MiningField name="rate" usageType="predicted"/>
                        <MiningField name="year" usageType="active"/>
                        <MiningField name="attribute1" usageType="active"/>
                    </MiningSchema>

                    <Output>
                        <OutputField name="Predicted_rate" feature="predictedValue"/>
                    </Output>
                    <RegressionTable intercept="5000">
                        <NumericPredictor name="year" exponent="1" coefficient="-0.75"/>
                    </RegressionTable>
                </RegressionModel>
            </Segment>
        </Segmentation>
    </MiningModel>
</PMML>

DynamicPmmlModelSelectionExample.java

package de.tutorials.training;

import org.dmg.pmml.FieldName;
import org.dmg.pmml.PMML;
import org.jpmml.evaluator.FieldValue;
import org.jpmml.evaluator.ModelEvaluator;
import org.jpmml.evaluator.ModelEvaluatorFactory;
import org.jpmml.manager.PMMLManager;
import org.jpmml.model.ImportFilter;
import org.jpmml.model.JAXBUtil;
import org.xml.sax.InputSource;

import java.io.InputStream;
import java.util.HashMap;
import java.util.LinkedHashMap;
import java.util.List;
import java.util.Map;

public class DynamicPmmlModelSelectionExample {

    public static void main(String[] args) throws Exception {

        String modelFile = "dynamic-model-selection-example.pmml.xml";
        InputStream inputStream = DynamicPmmlModelSelectionExample.class.getClassLoader().getResourceAsStream(modelFile);

        PMML pmml = JAXBUtil.unmarshalPMML(ImportFilter.apply(new InputSource(inputStream)));

        PMMLManager pmmlManager = new PMMLManager(pmml);

        ModelEvaluator<?> evaluator = (ModelEvaluator<?>) pmmlManager.getModelManager(ModelEvaluatorFactory.getInstance());

        System.out.println(score(evaluator, new HashMap<String,Object>(){
            {
                put("year",2014);
                put("attribute1","1330");
            }
        }));

        System.out.println(score(evaluator, new HashMap<String,Object>(){
            {
                put("year",2014);
                put("attribute1","1315");
            }
        }));

        System.out.println(score(evaluator, new HashMap<String,Object>(){
            {
                put("year",2014);
                put("attribute1","1345");
            }
        }));
    }

    private static Map<FieldName, ?> score(ModelEvaluator<?> evaluator, Map<String, Object> input) {
        
        Map<FieldName, FieldValue> arguments = new LinkedHashMap<FieldName, FieldValue>();

        List<FieldName> activeFields = evaluator.getActiveFields();
        for(FieldName activeField : activeFields){
        
            Object rawValue = input.get(activeField.getValue());

            FieldValue activeValue = evaluator.prepare(activeField, rawValue);

            arguments.put(activeField, activeValue);
        }

        return evaluator.evaluate(arguments);
    }
}

output.txt

{rate=18993.0, Predicted_rate=18993.0}
{rate=13021.0, Predicted_rate=13021.0}
{rate=3489.5, Predicted_rate=3489.5}