multiclass classification performance error?

ThisGuy

Hi, all. Performed a fairly straightforward classification by regression problem with SVM, and got a multiclass classification performance error. Problem is part of a grad class, so it's a shared, known problem set with a known outcome. At least others are not experiencing it, but I've received the errors on two different computers, even using the professor's XML code that worked for him. Any one else experience this error? Ideas on how to fix? Code below for reference.

ERROR CONTENT: "Error executing background job 'Calculating result: multiclass_classification_perfromance': java.lang.ArrayIndexOutOfBoundsException: 0"

CODE:

<参数键= "使用_except_expression" value="false"/>

<参数键= value =“use_value_type_exception false"/>

<参数键= "使用_block_type_exception" value="false"/>

<参数键= "使用_except_expression" value="false"/>

<参数键= value =“use_value_type_exception false"/>

<参数键= "使用_block_type_exception" value="false"/>

<参数键= "使用_local_random_seed" value="false"/>

<参数键= "使用_local_random_seed" value="false"/>

<参数键= "使用_example_weights" value="true"/>

<连接from_op = from_port = " averagabl“验证”e 1" to_port="result 3"/>

jwpfau

Hi,

thank you for the bug report.

Your "Filters Examples" is set to "Match all" which will result in 0 examples left since the "CTS state" can't be "Severe" and "Without" at the same time.

Either use "does not equals" to filter them out, or "Match any" to only include "Severe" and "Without" CTS states.

Greetings,
Jonas