Backward elimination, forward selection and optimize selection operators

searcher

I'm trying to use Backward elimination, forward selection and optimize selection operators to find useful predictors for scoring in my data.
I have 50 variables, 1500 samples.
How long will the Backward elimination, forward selection and optimize selection operators work for my data? Is any standards of efficiency  for this algorithms ?

TobiasMalbrecht

Hi,

searcher wrote:
I'm trying to use Backward elimination, forward selection and optimize selection operators to find useful predictors for scoring in my data. I have 50 variables, 1500 samples. How long will the Backward elimination, forward selection and optimize selection operators work for my data? Is any standards of efficiency  for this algorithms ?

that naturally depends on several factors, i.e. the computer you are working on, the learning scheme and the ordering of your attributes in combination with the parameter settings. Presumably, you will have to work (try) it out for yourself. Just start with a learning scheme that does not take very long (e.g. Naive Bayes) - assumed it works on your data.

Kind regards,
Tobias

searcher

I'm using computer with 7300 M Memory & 8 core processor with 2 Gh per core .
I'm using Logistic Regression in this way

<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<process version="5.0">
  <context>
    <input/>
    <output/>
    <macros/>
  </context>
  <operator activated="true" class="process" expanded="true" name="Process">
    <process expanded="true" height="-20" width="-50">
      <operator activated="true" class="optimize_selection_backward" expanded="true" height="94" name="Backward Elimination" width="90" x="150" y="48">
        <process expanded="true">
          <operator activated="true" class="split_validation" expanded="true" name="Validation">
            <parameter key="parallelize_training" value="true"/>
            <parameter key="parallelize_testing" value="true"/>
            <process expanded="true">
              <operator activated="false" class="multiply" expanded="true" name="Multiply"/>
              <operator activated="false" class="nominal_to_numerical" expanded="true" name="Nominal to Numerical (2)">
                <parameter key="create_view" value="true"/>
                <parameter key="attribute_filter_type" value="single"/>
                <parameter key="attribute" value="DEFSTATUS"/>
              </operator>
              <operator activated="false" class="polynomial_regression" expanded="true" name="Polynomial Regression">
                <parameter key="use_local_random_seed" value="true"/>
              </operator>
              <operator activated="true" class="multiply" expanded="true" name="Multiply (2)"/>
              <operator activated="true" class="logistic_regression" expanded="true" name="Logistic Regression">
                <parameter key="kernel_type" value="polynomial"/>
                <parameter key="convergence_epsilon" value="1.0E-5"/>
              </operator>
              <connect from_port="training" to_op="Multiply (2)" to_port="input"/>
              <connect from_op="Multiply (2)" from_port="output 1" to_port="through 1"/>
              <connect from_op="Multiply (2)" from_port="output 2" to_op="Logistic Regression" to_port="training set"/>
              <connect from_op="Logistic Regression" from_port="model" to_port="model"/>
              <portSpacing port="source_training" spacing="0"/>
              <portSpacing port="sink_model" spacing="0"/>
              <portSpacing port="sink_through 1" spacing="0"/>
              <portSpacing port="sink_through 2" spacing="0"/>
            </process>
            <process expanded="true">
              <operator activated="false" class="select_attributes" expanded="true" name="Select Attributes (5)">
                <parameter key="attribute_filter_type" value="subset"/>
                <parameter key="attributes" value="DEFSTATUS|AMOUNT_USD|MONTHLY_REPAYMENT_USD|TERM|DISB_FEE|MONTHLY_FEE|INTEREST_RATE|AGREEM*ENTER|OWNERSHIP*POST|ENTER"/>
                <parameter key="include_special_attributes" value="true"/>
              </operator>
              <operator activated="true" class="multiply" expanded="true" name="Multiply (3)"/>
              <operator activated="true" class="apply_model" expanded="true" name="Apply Model (3)">
                <list key="application_parameters"/>
              </operator>
              <operator activated="true" class="performance" expanded="true" name="Selection_Training"/>
              <operator activated="true" class="apply_model" expanded="true" name="Apply Model">
                <list key="application_parameters"/>
              </operator>
              <operator activated="true" class="performance" expanded="true" name="Test"/>
              <connect from_port="model" to_op="Multiply (3)" to_port="input"/>
              <connect from_port="test set" to_op="Apply Model" to_port="unlabelled data"/>
              <connect from_port="through 1" to_op="Apply Model (3)" to_port="unlabelled data"/>
              <connect from_op="Multiply (3)" from_port="output 1" to_op="Apply Model" to_port="model"/>
              <connect from_op="Multiply (3)" from_port="output 2" to_op="Apply Model (3)" to_port="model"/>
              <connect from_op="Apply Model (3)" from_port="labelled data" to_op="Selection_Training" to_port="labelled data"/>
              <connect from_op="Selection_Training" from_port="performance" to_port="averagable 2"/>
              <connect from_op="Apply Model" from_port="labelled data" to_op="Test" to_port="labelled data"/>
              <connect from_op="Test" from_port="performance" to_port="averagable 1"/>
              <portSpacing port="source_model" spacing="0"/>
              <portSpacing port="source_test set" spacing="0"/>
              <portSpacing port="source_through 1" spacing="0"/>
              <portSpacing port="source_through 2" spacing="0"/>
              <portSpacing port="sink_averagable 1" spacing="0"/>
              <portSpacing port="sink_averagable 2" spacing="0"/>
              <portSpacing port="sink_averagable 3" spacing="0"/>
            </process>
          </operator>
          <connect from_port="example set" to_op="Validation" to_port="training"/>
          <connect from_op="Validation" from_port="averagable 2" to_port="performance"/>
          <portSpacing port="source_example set" spacing="0"/>
          <portSpacing port="sink_performance" spacing="0"/>
        </process>
      </operator>
      <operator activated="true" class="split_validation" expanded="true" height="130" name="Validation (2)" width="90" x="378" y="111">
        <process expanded="true">
          <operator activated="false" class="multiply" expanded="true" name="Multiply (4)"/>
          <operator activated="false" class="nominal_to_numerical" expanded="true" name="Nominal to Numerical (5)">
            <parameter key="create_view" value="true"/>
            <parameter key="attribute_filter_type" value="single"/>
            <parameter key="attribute" value="DEFSTATUS"/>
          </operator>
          <operator activated="false" class="polynomial_regression" expanded="true" name="Polynomial Regression (2)">
            <parameter key="use_local_random_seed" value="true"/>
          </operator>
          <operator activated="true" class="multiply" expanded="true" name="Multiply (5)"/>
          <operator activated="true" class="logistic_regression" expanded="true" name="Logistic Regression (2)">
            <parameter key="kernel_type" value="polynomial"/>
            <parameter key="convergence_epsilon" value="1.0E-5"/>
          </operator>
          <connect from_port="training" to_op="Multiply (5)" to_port="input"/>
          <connect from_op="Multiply (5)" from_port="output 1" to_port="through 1"/>
          <connect from_op="Multiply (5)" from_port="output 2" to_op="Logistic Regression (2)" to_port="training set"/>
          <connect from_op="Logistic Regression (2)" from_port="model" to_port="model"/>
          <portSpacing port="source_training" spacing="0"/>
          <portSpacing port="sink_model" spacing="0"/>
          <portSpacing port="sink_through 1" spacing="0"/>
          <portSpacing port="sink_through 2" spacing="0"/>
        </process>
        <process expanded="true">
          <operator activated="false" class="select_attributes" expanded="true" name="Select Attributes (7)">
            <parameter key="attribute_filter_type" value="subset"/>
            <parameter key="attributes" value="DEFSTATUS|AMOUNT_USD|MONTHLY_REPAYMENT_USD|TERM|DISB_FEE|MONTHLY_FEE|INTEREST_RATE|AGREEM*ENTER|OWNERSHIP*POST|ENTER"/>
            <parameter key="include_special_attributes" value="true"/>
          </operator>
          <operator activated="true" class="multiply" expanded="true" name="Multiply (6)"/>
          <operator activated="true" class="apply_model" expanded="true" name="Apply Model (4)">
            <list key="application_parameters"/>
          </operator>
          <operator activated="true" class="performance" expanded="true" name="Training (2)"/>
          <operator activated="true" class="apply_model" expanded="true" name="Apply Model (5)">
            <list key="application_parameters"/>
          </operator>
          <operator activated="true" class="performance" expanded="true" name="Test (2)"/>
          <connect from_port="model" to_op="Multiply (6)" to_port="input"/>
          <connect from_port="test set" to_op="Apply Model (5)" to_port="unlabelled data"/>
          <connect from_port="through 1" to_op="Apply Model (4)" to_port="unlabelled data"/>
          <connect from_op="Multiply (6)" from_port="output 1" to_op="Apply Model (5)" to_port="model"/>
          <connect from_op="Multiply (6)" from_port="output 2" to_op="Apply Model (4)" to_port="model"/>
          <connect from_op="Apply Model (4)" from_port="labelled data" to_op="Training (2)" to_port="labelled data"/>
          <connect from_op="Training (2)" from_port="performance" to_port="averagable 2"/>
          <connect from_op="Apply Model (5)" from_port="labelled data" to_op="Test (2)" to_port="labelled data"/>
          <connect from_op="Test (2)" from_port="performance" to_port="averagable 1"/>
          <portSpacing port="source_model" spacing="0"/>
          <portSpacing port="source_test set" spacing="0"/>
          <portSpacing port="source_through 1" spacing="0"/>
          <portSpacing port="source_through 2" spacing="0"/>
          <portSpacing port="sink_averagable 1" spacing="0"/>
          <portSpacing port="sink_averagable 2" spacing="0"/>
          <portSpacing port="sink_averagable 3" spacing="0"/>
        </process>
      </operator>
      <connect from_port="input 1" to_op="Backward Elimination" to_port="example set"/>
      <connect from_op="Backward Elimination" from_port="example set" to_op="Validation (2)" to_port="training"/>
      <connect from_op="Backward Elimination" from_port="attribute weights" to_port="result 1"/>
      <connect from_op="Backward Elimination" from_port="performance" to_port="result 2"/>
      <connect from_op="Validation (2)" from_port="model" to_port="result 5"/>
      <connect from_op="Validation (2)" from_port="averagable 1" to_port="result 3"/>
      <connect from_op="Validation (2)" from_port="averagable 2" to_port="result 4"/>
      <portSpacing port="source_input 1" spacing="0"/>
      <portSpacing port="source_input 2" spacing="0"/>
      <portSpacing port="sink_result 1" spacing="0"/>
      <portSpacing port="sink_result 2" spacing="0"/>
      <portSpacing port="sink_result 3" spacing="0"/>
      <portSpacing port="sink_result 4" spacing="0"/>
      <portSpacing port="sink_result 5" spacing="0"/>
      <portSpacing port="sink_result 6" spacing="0"/>
    </process>
  </operator>
</process>

Is this the right way ?
Thanks, I'll try to use Naive Bayes.
For ending of algorithm with Log Regression i waiting already almost 6 days/

haddock

Hi,

Why not post some code that works?  

searcher

I'm sorry for that, but i cant post data I'm using in scoring.
I just wanted to show the structure and ask is this right using of backward selection operator.

haddock

Hi,

Then the short answer is that your process does not do what you describe as your goal, whatever the data.

searcher

Could you suggest  the way of using backward elimination, forward selection and optimize selection operators to find most useful predictors in scoring?

haddock

Hi,

Indeed I could, but as paid consultancy, but you would be better off going to RM staffers, and that would help the cause more.

searcher

searcher wrote:
Hi, Then the short answer is that your process does not do what you describe as your goal, whatever the data.

Ok.So, why do you think that the way of using backward elimination i posted is wrong?