Options

Forcing YAGGA2 to use an attribute in every individual causes senseless WEPTrees

BAMBAMBAMBAMBAMBAM Member Posts: 20 Maven
edited November 2018 in Help
Hi,

I've got a YAGGA2 process that feeds attributes to a WEPTree, and I would like the WEPTree to always use a particular attribute (Entry_OVR_Close).  I've come up with a process to do this, but it yields very strange, contradictory WEPTrees.  Here is a sample output:

REPTree
============

Entry_OVR_Close < 0
|  Entry_OVR_Close < 0 : 0.11 (264/0.1) [139/0.02]
|  Entry_OVR_Close >= 0 : 0.03 (587/0.02) [316/0.02]
Entry_OVR_Close >= 0
|  Entry_OVR_Close < 0
|  |  Entry_OVR_Close < 0 : -0.02 (590/0.01) [256/0.01]
|  |  Entry_OVR_Close >= 0 : -0.05 (221/0.02) [118/0.02]
|  Entry_OVR_Close >= 0 : -0.1 (470/0.01) [238/0.03]

For this particular output, the WEPTree could choose from roughly eight possible attributes, but ended up choosing only Entry_OVR_Close.  The weird thing is that the rules don't make sense - each leaf covers the possibilities that both Entry_OVR_Close<0 AND Entry_OVR_Close>=0.  Any ideas on how I could prevent this?

Here is my process:
<operator name="Root" class="Process" expanded="yes">
    <description text="TYPE #ylt# 0.5 |  Entry_OVR_Close #ylt# 0.5 |  |  PercentR_50 #ylt# 0.5 |  |  |  PercentR_20 #ylt# 0.5 |  |  |  |  prcK_201 #ylt# 0.5 |  |  |  |  |  prcK_141 #ylt# 0.5 |  |  |  |  |  |  prcK_341 #ylt# 0.5 |  |  |  |  |  |  |  PercentR_8 #ylt# 0.5 |  |  |  |  |  |  |  |  RSI_8C #ylt# 0.5 : 0.01 (13126/0) [6575/0] |  |  |  |  |  |  |  |  RSI_8C #ygt#= 0.5 : 0.03 (526/0) [275/0] |  |  |  |  |  |  |  PercentR_8 #ygt#= 0.5 : 0.04 (576/0) [296/0] |  |  |  |  |  |  prcK_341 #ygt#= 0.5 : 0.08 (970/0) [516/0] |  |  |  |  |  prcK_141 #ygt#= 0.5 : 0.08 (1005/0) [517/0] |  |  |  |  prcK_201 #ygt#= 0.5 : 0.08 (1309/0) [665/0] |  |  |  PercentR_20 #ygt#= 0.5 : 0.08 (1405/0) [726/0] |  |  PercentR_50 #ygt#= 0.5 : 0.08 (1315/0) [642/0] |  Entry_OVR_Close #ygt#= 0.5 : 0.13 (5026/0.01) [2498/0.01] TYPE #ygt#= 0.5 |  Entry_OVR_Close #ylt# 0.5 |  |  PercentR_20 #ylt# 0.5 : 0.14 (2934/0) [1406/0] |  |  PercentR_20 #ygt#= 0.5 |  |  |  SYM_DIF_SPY_ArnUp8 #ylt# 0.5 : 0.16 (5992/0) [2909/0] |  |  |  SYM_DIF_SPY_ArnUp8 #ygt#= 0.5 : 0.2 (665/0) [364/0] |  Entry_OVR_Close #ygt#= 0.5 |  |  CCI_50C #ylt# 0.5 : 0.31 (2032/0.05) [1082/0.05] |  |  CCI_50C #ygt#= 0.5 : 0.48 (583/0) [262/0]"/>
    <parameter key="logverbosity" value="error"/>
    <operator name="LoadData" class="OperatorChain" expanded="yes">
        <operator name="MacroDefinition" class="MacroDefinition">
            <list key="macros">
              <parameter key="baseName" value="longs"/>
            </list>
        </operator>
        <operator name="ExampleSource" class="ExampleSource" activated="no">
            <parameter key="attributes" value="jpriorsTrades.att"/>
            <parameter key="local_random_seed" value="2001"/>
        </operator>
        <operator name="CSVExampleSource" class="CSVExampleSource">
            <parameter key="filename" value="I:\Trading\RapidMiner\MetaTrader\longs.csv"/>
            <parameter key="label_name" value="EFF"/>
            <parameter key="id_name" value="id"/>
        </operator>
        <operator name="ExampleSetWriter" class="ExampleSetWriter">
            <parameter key="example_set_file" value="longs.rmd"/>
            <parameter key="attribute_description_file" value="longs.att"/>
        </operator>
        <operator name="ChangeAttributeRole" class="ChangeAttributeRole" activated="no">
            <parameter key="name" value="TYPE"/>
            <parameter key="target_role" value="i1"/>
        </operator>
        <operator name="ChangeAttributeRole (2)" class="ChangeAttributeRole">
            <parameter key="name" value="Entry_OVR_Close"/>
            <parameter key="target_role" value="i2"/>
        </operator>
        <operator name="ChangeAttributeRole (3)" class="ChangeAttributeRole" activated="no">
            <parameter key="name" value="CCI_50C"/>
            <parameter key="target_role" value="i3"/>
        </operator>
    </operator>
    <operator name="YAGGA2" class="YAGGA2" breakpoints="after" expanded="yes">
        <parameter key="local_random_seed" value="2001"/>
        <parameter key="population_size" value="1000"/>
        <parameter key="maximum_number_of_generations" value="25"/>
        <parameter key="generations_without_improval" value="3"/>
        <parameter key="keep_best_individual" value="true"/>
        <parameter key="p_initialize" value="0.02"/>
        <parameter key="use_plus" value="false"/>
        <parameter key="use_diff" value="true"/>
        <parameter key="use_div" value="true"/>
        <parameter key="p_mutation" value="0.1"/>
        <parameter key="use_square_roots" value="true"/>
        <parameter key="use_sin" value="false"/>
        <parameter key="use_exp" value="false"/>
        <parameter key="use_absolute_values" value="false"/>
        <parameter key="equivalence_use_statistics" value="false"/>
        <parameter key="constant_generation_prob" value="0.0"/>
        <operator name="XValidation" class="XValidation" expanded="yes">
            <parameter key="keep_example_set" value="true"/>
            <parameter key="number_of_validations" value="3"/>
            <parameter key="sampling_type" value="shuffled sampling"/>
            <parameter key="local_random_seed" value="1992"/>
            <operator name="OperatorChain" class="OperatorChain" expanded="yes">
                <operator name="ChangeAttributeRole (4)" class="ChangeAttributeRole" activated="no">
                    <parameter key="name" value="TYPE"/>
                </operator>
                <operator name="ChangeAttributeRole (6)" class="ChangeAttributeRole">
                    <parameter key="name" value="Entry_OVR_Close"/>
                </operator>
                <operator name="ChangeAttributeRole (5)" class="ChangeAttributeRole" activated="no">
                    <parameter key="name" value="CCI_50C"/>
                </operator>
                <operator name="Construction2Names" class="Construction2Names">
                </operator>
                <operator name="W-REPTree" class="W-REPTree">
                    <parameter key="keep_example_set" value="true"/>
                    <parameter key="M" value="100.0"/>
                </operator>
                <operator name="LinearRegression" class="LinearRegression" activated="no">
                    <parameter key="keep_example_set" value="true"/>
                </operator>
                <operator name="MemoryCleanUp" class="MemoryCleanUp">
                </operator>
            </operator>
            <operator name="ApplierChain" class="OperatorChain" expanded="yes">
                <operator name="Applier" class="ModelApplier">
                    <parameter key="keep_model" value="true"/>
                    <list key="application_parameters">
                    </list>
                </operator>
                <operator name="RegressionPerformance" class="RegressionPerformance">
                    <parameter key="main_criterion" value="spearman_rho"/>
                    <parameter key="spearman_rho" value="true"/>
                    <parameter key="use_example_weights" value="false"/>
                </operator>
                <operator name="PerfWriter" class="ResultWriter">
                    <parameter key="result_file" value="%{baseName}_attrWeights.txt"/>
                </operator>
                <operator name="ProcessLog" class="ProcessLog">
                    <list key="log">
                      <parameter key="Perf" value="operator.RegressionPerformance.value.performance"/>
                      <parameter key="Tries" value="operator.RegressionPerformance.value.applycount"/>
                      <parameter key="len" value="operator.YAGGA2.value.average_length"/>
                      <parameter key="gen" value="operator.YAGGA2.value.generation"/>
                    </list>
                </operator>
                <operator name="ChangeAttributeRole (7)" class="ChangeAttributeRole" activated="no">
                    <parameter key="name" value="TYPE"/>
                    <parameter key="target_role" value="i1"/>
                </operator>
                <operator name="ChangeAttributeRole (8)" class="ChangeAttributeRole">
                    <parameter key="name" value="Entry_OVR_Close"/>
                    <parameter key="target_role" value="i2"/>
                </operator>
                <operator name="ChangeAttributeRole (9)" class="ChangeAttributeRole" activated="no">
                    <parameter key="name" value="CCI_50C"/>
                    <parameter key="target_role" value="i3"/>
                </operator>
            </operator>
        </operator>
    </operator>
    <operator name="AttributeWeightSelection" class="AttributeWeightSelection" breakpoints="after">
        <parameter key="weight" value="0.0"/>
        <parameter key="weight_relation" value="greater"/>
    </operator>
    <operator name="FinalTree" class="W-REPTree" breakpoints="after">
        <parameter key="keep_example_set" value="true"/>
        <parameter key="M" value="50.0"/>
    </operator>
    <operator name="FinalRegression" class="LinearRegression" breakpoints="after" activated="no">
        <parameter key="keep_example_set" value="true"/>
    </operator>
    <operator name="ModelWriter" class="ModelWriter" breakpoints="after">
        <parameter key="model_file" value="LinearRegressionModel.xml"/>
        <parameter key="output_type" value="XML"/>
    </operator>
    <operator name="ModelApplier" class="ModelApplier" breakpoints="after">
        <list key="application_parameters">
        </list>
    </operator>
    <operator name="FinalPerf" class="RegressionPerformance" breakpoints="after">
        <parameter key="main_criterion" value="spearman_rho"/>
        <parameter key="spearman_rho" value="true"/>
        <parameter key="use_example_weights" value="false"/>
    </operator>
</operator>

thanks in advance,
John

Answers

  • Options
    landland RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 2,531 Unicorn
    Hi,
    sorry I cannot say anything about the W-REPTree because it's not one of our learners and I don't have a clue, what it does anyway...

    Greetings,
      Sebastian
  • Options
    BAMBAMBAMBAMBAMBAM Member Posts: 20 Maven
    Sebastion, could you suggest a learner that Rapid does know about, which outputs human-understandable rules and can handle numeric inputs and outputs?

    thank you,
    JOhn
  • Options
    landland RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 2,531 Unicorn
    Hi,
    unfortunately our own regression tree was dropped during one of the early version changes in 4.x. So unfortunately I cannot recommend you anyone.

    Greetings,
      Sebastian
Sign In or Register to comment.