Independent fields, also called input fields or features, are the variables used to make predictions. 
In PMML these are typically defined in the DataDictionary and MiningSchema elements.

The dependent field is the output or target variable that the model predicts. 
It's also defined in both the DataDictionary and MiningSchema, but with usageType="predicted"

independent (ie. input) fields
独立（即输入）字段



Key Differences

Usage Type:
Independent: usageType="active"
Dependent: usageType="predicted"

Role:
Independent fields are inputs to the model
Dependent field is the output of the model


Multiple vs Single:
There can be multiple independent fields
Typically only one dependent field (though some models support multiple outputs)





<?xml version="1.0" encoding="UTF-8"?>
<PMML xmlns="http://www.dmg.org/PMML-4_4" version="4.4">
    <Header copyright="Example Corp" description="Simple Linear Regression Model">
        <Application name="Example App" version="1.0"/>
    </Header>
    <DataDictionary numberOfFields="3">
        <DataField name="x1" optype="continuous" dataType="double"/>
        <DataField name="x2" optype="continuous" dataType="double"/>
        <DataField name="y" optype="continuous" dataType="double"/>
    </DataDictionary>
    <RegressionModel modelName="SimpleLinearRegression" functionName="regression">
        <MiningSchema>
            <MiningField name="x1" usageType="active"/>
            <MiningField name="x2" usageType="active"/>
            <MiningField name="y" usageType="predicted"/>
        </MiningSchema>
        <RegressionTable intercept="1.0">
            <NumericPredictor name="x1" exponent="1" coefficient="0.5"/>
            <NumericPredictor name="x2" exponent="1" coefficient="0.3"/>
        </RegressionTable>
    </RegressionModel>
</PMML>