I clustered some kinds of flowers into 3 groups. E.g cluster 1 has 2 kinds of flowers,cluster 2 has 15 kinds of flowers and the last cluster has 18 kinds of flowers. Now i want to predict the output variable which is common in all of flowers and it is continuous numerical number with neural network. I have 7 input variables and 1 output variable with 3 levels( of 3 kinds of flowers on clusters) . how can i encode the output variable that i can use "nnet" function in R? I mean, if i want to use nnet function and the output called " m" how should i encode it to be able to write the code below in R:
nnet(m~x1+x2+x3+x4+x5+x6+x7,data=train.data, size=9)
Any help will be appreciated.
Hello Ghazaleh,
You may have already seen it, but this is a good place to start:
https://beckmw.wordpress.com/tag/nnet/
There are many examples, but specifically in response to your question, there are several examples where the author demonstrates setting up multi-level response variables for nnet.
Essentially, you should set up a dataframe with your multilevel response as columns (resp).
Next, combine your independent variables, as columns (ind.vars), and dependent variables:
dat.in
If you found my answer useful, please vote it up! Thanks, and good luck!
- Badges
- Science topic