topic Re: Categorical data analysis in Statistical Procedures

Categorical data analysis

bhadrappa_molgi — Wed, 23 Mar 2016 15:03:45 GMT

I have 5 categorical variables like client, location, types of services, prodcucts , bill of materials used etc. Can I use multinomial logistic regression for predicting the target variable which is categorical which more than 2 levels with other 5-6 categorical variables which are multilevel too?

can you suggest a good technique to handle this, I have more than 500000 rows of data , so I want to predict the on the future data. and also there could seasonality effect inside the data.

Re: Categorical data analysis

Ksharp — Thu, 24 Mar 2016 06:51:17 GMT

"Can I use multinomial logistic regression for predicting the target variable which is categorical which more than 2 levels "
Answer is yes. That lead to General Logistic Model.

An alternative method is using Decision Tree which Data Scientist love it . But you need SAS/EM to run it.


"so I want to predict the on the future data. and also there could seasonality effect inside the data."
Do you mean these data are Time Series data ? then you need SAS/ETS .