"Logistic Regression cannot handle polynominal label."

arunpushkar

i am using "read Database" operator to read training data from Mysql having 3 integer attributes and one varchar(4) attribute which will act as "lable" for classification and takes value 'yes' or 'no'. now when i use "logistic regression" for training model i get following error"Logistic Regression cannot handle polynomial label" but if i read csv data from csv file and set lable attribute to "nominal" it work's properly. but when i read data from database it gives error, first of all how to make any attribute read from database "nominal" so that it can be used by "Logical Regression". in fact i used "text to nominal" operator but it didn't work.
XML code for it as follows:

<宏/ >










<枚举关键= "parameters"/>












<参数键= "查询" value = " SELECT * & # 10;从“scoringData`"/>
<枚举关键= "parameters"/>

Skirzynski

Indeed it is strange that you get different behavior for the same data. Please check if the "Read CSV" operator really returns a nominal and not a binominal label. Unfortunately I do not have your CSV file nor your database to confirm your observation.

To solve your problem: Convert the type of your label to binominal explicitly. For this you can use the "Nominal to Binominal" operator.

<宏/ >










<枚举关键= "parameters"/>
















<参数键= "查询" value = " SELECT * & # 10;从“scoringData`"/>
<枚举关键= "parameters"/>