"Predictive analysis - problems with accuracy."

Hi contributors,

First of all, this is my third message and I really appreciate the effort of the experts. Keep up the good work!

Now after a few lessons of RM I got finally (in my opinion) something like a good design/model. Let me quick describe the model: Two sets, training and testing. I have used a windowing operator to devide the data in to blocks of 5days before today and which have to predict the value of tomorrow.

With a sliding window validation operator, NN network and a performance operator I would like to get the best accuracy I can get. Below in my XML you could see the model.

Now my question is: How could I increase my accuracy because this is really low. I would like to get a higher accuracy of the prediction of tomorrow. It seems that you have to play with the parameters but I'm not understanding all of them. In my opinion the next parameters could have some influence on the accuracy.

- Windowing Operator - Windowsize, stepsize, horizon.
I have set the windowsize on 5 because I think information about 5 days is enough but I could easily change this. Stepsize is 1 because you would like to predict every day and horizon of 2 because you predict the value of tomorrow and not today.

- Validation operator - Training window width, training window step size, test window with and hoirzon.
- Forecasting Performance - Horizon.

Either for this parameter i don't understand this meaning.

I could also play with the training block in the validation operator. For example NN-network, linear regression or SVM. Do someone sugest something else after seeing my problem and dataset?

Hopefully I'm clear enough. If the problem/questions is/are not clear please leave a comment!

Thanks for helping!

Maurits

I really don't understand these parameters!?

Below my XML code:

<?xml version="1.0" encoding="UTF-8"?><process version="7.6.001">

</context>

</operator>

</operator>

</operator>

</process>

</operator>

</operator>

</process>

</operator>

</operator>

</operator>

</operator>

</process>

</operator>

</process>

Find more posts tagged with

AI Studio

Performance

Getting Started

Accepted answers

All comments

Thomas_Ott

What you can do now is use an Optimize Parameter operator to tune your NN parameters like Momentum, Learning Rate, and Training Cycles. It will go through all those parameters and deliver the optimized model that you can use downstream in the process. See the sample process attached below.

<?xml version="1.0" encoding="UTF-8"?><process version="7.6.002">
  <context>
    <input/>
    <output/>
    <macros/>
  </context>
  <operator activated="true" class="process" compatibility="7.6.002" expanded="true" name="Process">
    <process expanded="true">
      <operator activated="true" class="retrieve" compatibility="7.6.002" expanded="true" height="68" name="Retrieve test data only flow oktober days train set" width="90" x="45" y="34">
        <parameter key="repository_entry" value="//Local Repository/data/test data only flow oktober days train set"/>
      </operator>
      <operator activated="true" class="series:windowing" compatibility="7.4.000" expanded="true" height="82" name="Windowing" width="90" x="179" y="34">
        <parameter key="window_size" value="5"/>
        <parameter key="create_label" value="true"/>
        <parameter key="label_attribute" value="A"/>
        <parameter key="horizon" value="2"/>
      </operator>
      <operator activated="true" class="optimize_parameters_grid" compatibility="7.6.002" expanded="true" height="124" name="Optimize Parameters (Grid)" width="90" x="313" y="34">
        <list key="parameters">
          <parameter key="Neural Net.training_cycles" value="[100;1000;9;linear]"/>
          <parameter key="Neural Net.learning_rate" value="[0.1;0.9;8;linear]"/>
          <parameter key="Neural Net.momentum" value="[0.1;0.9;8;linear]"/>
        </list>
        <process expanded="true">
          <operator activated="true" class="series:sliding_window_validation" compatibility="7.4.000" expanded="true" height="124" name="Validation" width="90" x="112" y="34">
            <parameter key="training_window_width" value="20"/>
            <parameter key="training_window_step_size" value="5"/>
            <parameter key="test_window_width" value="10"/>
            <process expanded="true">
              <operator activated="true" class="neural_net" compatibility="7.6.002" expanded="true" height="82" name="Neural Net" width="90" x="179" y="34">
                <list key="hidden_layers"/>
              </operator>
              <connect from_port="training" to_op="Neural Net" to_port="training set"/>
              <connect from_op="Neural Net" 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"/>
            </process>
            <process expanded="true">
              <operator activated="true" class="apply_model" compatibility="7.6.002" expanded="true" height="82" name="Apply Model" width="90" x="45" y="34">
                <list key="application_parameters"/>
              </operator>
              <operator activated="true" class="series:forecasting_performance" compatibility="7.4.000" expanded="true" height="82" name="Performance" width="90" x="179" y="34">
                <parameter key="horizon" value="2"/>
              </operator>
              <connect from_port="model" to_op="Apply Model" to_port="model"/>
              <connect from_port="test set" to_op="Apply Model" to_port="unlabelled data"/>
              <connect from_op="Apply Model" from_port="labelled data" to_op="Performance" to_port="labelled data"/>
              <connect from_op="Performance" 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="sink_averagable 1" spacing="0"/>
              <portSpacing port="sink_averagable 2" spacing="0"/>
            </process>
          </operator>
          <connect from_port="input 1" to_op="Validation" to_port="training"/>
          <connect from_op="Validation" from_port="model" to_port="result 1"/>
          <connect from_op="Validation" from_port="averagable 1" to_port="performance"/>
          <portSpacing port="source_input 1" spacing="0"/>
          <portSpacing port="source_input 2" spacing="0"/>
          <portSpacing port="sink_performance" spacing="0"/>
          <portSpacing port="sink_result 1" spacing="0"/>
          <portSpacing port="sink_result 2" spacing="0"/>
        </process>
      </operator>
      <operator activated="true" class="retrieve" compatibility="7.6.002" expanded="true" height="68" name="Retrieve test data only flow oktober days test set" width="90" x="45" y="340">
        <parameter key="repository_entry" value="//Local Repository/data/test data only flow oktober days test set"/>
      </operator>
      <operator activated="true" class="series:windowing" compatibility="7.4.000" expanded="true" height="82" name="Windowing (2)" width="90" x="179" y="340">
        <parameter key="window_size" value="5"/>
        <parameter key="create_label" value="true"/>
        <parameter key="label_attribute" value="A"/>
        <parameter key="horizon" value="2"/>
      </operator>
      <operator activated="true" class="apply_model" compatibility="7.6.002" expanded="true" height="82" name="Apply Model (2)" width="90" x="648" y="238">
        <list key="application_parameters"/>
      </operator>
      <connect from_op="Retrieve test data only flow oktober days train set" from_port="output" to_op="Windowing" to_port="example set input"/>
      <connect from_op="Windowing" from_port="example set output" to_op="Optimize Parameters (Grid)" to_port="input 1"/>
      <connect from_op="Optimize Parameters (Grid)" from_port="performance" to_port="result 3"/>
      <connect from_op="Optimize Parameters (Grid)" from_port="parameter" to_port="result 4"/>
      <connect from_op="Optimize Parameters (Grid)" from_port="result 1" to_op="Apply Model (2)" to_port="model"/>
      <connect from_op="Retrieve test data only flow oktober days test set" from_port="output" to_op="Windowing (2)" to_port="example set input"/>
      <connect from_op="Windowing (2)" from_port="example set output" to_op="Apply Model (2)" to_port="unlabelled data"/>
      <connect from_op="Apply Model (2)" from_port="labelled data" to_port="result 1"/>
      <connect from_op="Apply Model (2)" from_port="model" to_port="result 2"/>
      <portSpacing port="source_input 1" 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"/>
    </process>
  </operator>
</process>

SGolbert

In your current process I don't much sense on doing CV, because you already have an independent test dataset.

However, the CV makes sense inside the Optimize Parameters operator. It is nice to test several models using only the training dataset and then measure the performance of the final model in an independent dataset (sometimes called validation dataset).

maurits_freriks

Sorry i try to figure out what the meaning of "CV" is, but I couldn't find it.

MartinLiebig

@maurits_freriks,

it's Cross Validation (or X-Validation)

Best,

maurits_freriks

@Thomas_Ott @SGolbert

Alright clear enough, I will use the optimize parameter. Does this operator also change my sliding window validation parameter?

When I run your example test @Thomas_Ott, I got the error message: "Cannot reset network to a smaller learning rate".

I deleted 0-values in my trainig and testset, people on the forums said this might be a solution, but still got the error message. Do you know what I'm doing wrong?

Thomas_Ott

@maurits_freriks you might need to change the range of the Learning Rate to say 0.2 to 0.9 in optimization, it can't converge on the global minimum with that small of LR.

maurits_freriks

@Thomas_Ott

I've changed this to 0.9 but still the same error ..

<?xml version="1.0" encoding="UTF-8"?><process version="7.6.001">
<context>
<input/>
<output/>
<macros/>
</context>
<operator activated="true" class="process" compatibility="7.6.001" expanded="true" name="Process">
<process expanded="true">
<operator activated="true" class="retrieve" compatibility="7.6.001" expanded="true" height="68" name="Retrieve test data only flow oktober days train set.2" width="90" x="45" y="34">
<parameter key="repository_entry" value="//Local Repository/data/test data only flow oktober days train set.2"/>
</operator>
<operator activated="true" class="series:windowing" compatibility="7.4.000" expanded="true" height="82" name="Windowing" width="90" x="179" y="34">
<parameter key="window_size" value="5"/>
<parameter key="create_label" value="true"/>
<parameter key="label_attribute" value="A"/>
<parameter key="horizon" value="2"/>
</operator>
<operator activated="true" class="optimize_parameters_grid" compatibility="7.6.001" expanded="true" height="124" name="Optimize Parameters (Grid)" width="90" x="313" y="34">
<list key="parameters">
<parameter key="Neural Net.training_cycles" value="[100;1000;9;linear]"/>
<parameter key="Neural Net.learning_rate" value="[0.1;0.9;8;linear]"/>
<parameter key="Neural Net.momentum" value="[0.1;0.9;8;linear]"/>
</list>
<process expanded="true">
<operator activated="true" class="series:sliding_window_validation" compatibility="7.4.000" expanded="true" height="124" name="Validation" width="90" x="112" y="34">
<parameter key="training_window_width" value="20"/>
<parameter key="training_window_step_size" value="5"/>
<parameter key="test_window_width" value="10"/>
<process expanded="true">
<operator activated="true" class="neural_net" compatibility="7.6.001" expanded="true" height="82" name="Neural Net" width="90" x="179" y="34">
<list key="hidden_layers"/>
<parameter key="training_cycles" value="800"/>
<parameter key="learning_rate" value="0.9"/>
<parameter key="momentum" value="0.5"/>
</operator>
<connect from_port="training" to_op="Neural Net" to_port="training set"/>
<connect from_op="Neural Net" 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"/>
</process>
<process expanded="true">
<operator activated="true" class="apply_model" compatibility="7.6.001" expanded="true" height="82" name="Apply Model" width="90" x="45" y="34">
<list key="application_parameters"/>
</operator>
<operator activated="true" class="series:forecasting_performance" compatibility="7.4.000" expanded="true" height="82" name="Performance" width="90" x="179" y="34">
<parameter key="horizon" value="2"/>
</operator>
<connect from_port="model" to_op="Apply Model" to_port="model"/>
<connect from_port="test set" to_op="Apply Model" to_port="unlabelled data"/>
<connect from_op="Apply Model" from_port="labelled data" to_op="Performance" to_port="labelled data"/>
<connect from_op="Performance" 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="sink_averagable 1" spacing="0"/>
<portSpacing port="sink_averagable 2" spacing="0"/>
</process>
</operator>
<connect from_port="input 1" to_op="Validation" to_port="training"/>
<connect from_op="Validation" from_port="model" to_port="result 1"/>
<connect from_op="Validation" from_port="averagable 1" to_port="performance"/>
<portSpacing port="source_input 1" spacing="0"/>
<portSpacing port="source_input 2" spacing="0"/>
<portSpacing port="sink_performance" spacing="0"/>
<portSpacing port="sink_result 1" spacing="0"/>
<portSpacing port="sink_result 2" spacing="0"/>
</process>
</operator>
<operator activated="true" class="retrieve" compatibility="7.6.001" expanded="true" height="68" name="Retrieve test data only flow oktober days train set.2 (2)" width="90" x="45" y="340">
<parameter key="repository_entry" value="//Local Repository/data/test data only flow oktober days train set.2"/>
</operator>
<operator activated="true" class="series:windowing" compatibility="7.4.000" expanded="true" height="82" name="Windowing (2)" width="90" x="179" y="340">
<parameter key="window_size" value="5"/>
<parameter key="create_label" value="true"/>
<parameter key="label_attribute" value="A"/>
<parameter key="horizon" value="2"/>
</operator>
<operator activated="true" class="apply_model" compatibility="7.6.001" expanded="true" height="82" name="Apply Model (2)" width="90" x="648" y="238">
<list key="application_parameters"/>
</operator>
<connect from_op="Retrieve test data only flow oktober days train set.2" from_port="output" to_op="Windowing" to_port="example set input"/>
<connect from_op="Windowing" from_port="example set output" to_op="Optimize Parameters (Grid)" to_port="input 1"/>
<connect from_op="Optimize Parameters (Grid)" from_port="performance" to_port="result 3"/>
<connect from_op="Optimize Parameters (Grid)" from_port="parameter" to_port="result 4"/>
<connect from_op="Optimize Parameters (Grid)" from_port="result 1" to_op="Apply Model (2)" to_port="model"/>
<connect from_op="Retrieve test data only flow oktober days train set.2 (2)" from_port="output" to_op="Windowing (2)" to_port="example set input"/>
<connect from_op="Windowing (2)" from_port="example set output" to_op="Apply Model (2)" to_port="unlabelled data"/>
<connect from_op="Apply Model (2)" from_port="labelled data" to_port="result 1"/>
<connect from_op="Apply Model (2)" from_port="model" to_port="result 2"/>
<portSpacing port="source_input 1" 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"/>
</process>
</operator>
</process>

Thomas_Ott

@maurits_freriks do you have a sample data set I can use? I was saying you should change the lower value of LR to 0.2 or 0.25

abbasi_samira

Hello
I'm having trouble calculating accuracy
The precision of my decision tree was 72 before the new Windows installation
But now that I've changed Windows and installed the software, the accuracy is 60
Steps to achieve accuracy are the same
I do not know where the problem is
Now weka software has a precision of 72, but the RapidMiner software is 60
And I want to work with RapidMiner software
Where do you think the problem is?

MartinLiebig

Hi,

did you upgrade from 7.6 to 8.0 in the meanwhile?

if yes: We changed a bit in the decision trees. But that should not be worsen but increase your perfromance. You could try to set the compatibility level of your DecTree back to 7.6.

Cheers,

Martin