[R] Weighted polytomous logistic regression
yhwang127 at yahoo.com
Fri Sep 19 20:21:13 CEST 2008
I am trying to fit a polytomous logistic regression with inverted sampling probabity as weight. I've tried multinom(nnet) function. While the results look not right. For example, the results were not the same for the following two model. Should I normalize the weights first? Or there were some other functions to use?
More information about the R-help