Decision Tree

Rawan

Hi everyone,
After building decision tree how can I know

1. Tree depth
2. How many leaves (decision nodes) does it have

Thnx

BalazsBarany

Hi@Rawan,

check out Tree to Rules (included in RapidMiner) or "Decision Tree to Example Set" in the Converters extension.
你可以很容易地得到f决定节点的数量rom the number of rules.
The tree depth is a bit harder. Essentially, the number of & signs in the description of the rule (with Decision Tree to Example Set) is the number of decisions that were applied.

Here's an example using the Converters extension:

<参数键= " logverbosity " value = " init " / >
<帕拉meter key="random_seed" value="-1"/>
<帕拉meter key="send_mail" value="never"/>
<帕拉meter key="notification_email" value=""/>
<帕拉meter key="process_duration_for_mail" value="30"/>
<帕拉meter key="encoding" value="SYSTEM"/>


<帕拉meter key="repository_entry" value="//Samples/data/Iris"/>


<帕拉meter key="criterion" value="gain_ratio"/>
<帕拉meter key="maximal_depth" value="10"/>
<帕拉meter key="apply_pruning" value="true"/>
<帕拉meter key="confidence" value="0.1"/>
<帕拉meter key="apply_prepruning" value="true"/>
<帕拉meter key="minimal_gain" value="0.01"/>
<帕拉meter key="minimal_leaf_size" value="2"/>
<帕拉meter key="minimal_size_for_split" value="4"/>
<帕拉meter key="number_of_prepruning_alternatives" value="3"/>



<帕拉meter key="macro" value="nodes"/>
<帕拉meter key="macro_type" value="number_of_examples"/>
<帕拉meter key="statistics" value="average"/>
<帕拉meter key="attribute_name" value=""/>




<帕拉meter key="nodes" value="eval(%{nodes})"/>
<帕拉meter key="depth" value="length(replaceAll(Condition, "[^&]", "")) + 1"/>

<帕拉meter key="keep_all" value="true"/>


<帕拉meter key="use_default_aggregation" value="false"/>
<帕拉meter key="attribute_filter_type" value="all"/>
<帕拉meter key="attribute" value=""/>
<帕拉meter key="attributes" value=""/>
<帕拉meter key="use_except_expression" value="false"/>
<帕拉meter key="value_type" value="attribute_value"/>
<帕拉meter key="use_value_type_exception" value="false"/>
<帕拉meter key="except_value_type" value="time"/>
<帕拉meter key="block_type" value="attribute_block"/>
<帕拉meter key="use_block_type_exception" value="false"/>
<帕拉meter key="except_block_type" value="value_matrix_row_start"/>
<帕拉meter key="invert_selection" value="false"/>
<帕拉meter key="include_special_attributes" value="false"/>
<帕拉meter key="default_aggregation_function" value="average"/>

<帕拉meter key="depth" value="maximum"/>

<帕拉meter key="group_by_attributes" value="nodes"/>
<帕拉meter key="count_all_combinations" value="false"/>
<帕拉meter key="only_distinct" value="false"/>
<帕拉meter key="ignore_missings" value="true"/>

Regards,
Balázs

Rawan

BalazsBarany
sorry but i couldn't find & signs after I used the tree to rules to find the tree depth

BalazsBarany

Hi@Rawan,

that operator has a different output. There's no easy way to process the model description in RapidMiner, but you can copy and process it with an external tool like an editor. Tree to Rules separates the rules with "and".

The process I put into the discussion (you can copy the XML into RapidMiner Studio in the XML panel) uses the Converters extension. This contains the helpful operator that converts the tree to an example set which you can analyze inside RapidMiner.

Regards,
Balázs