Take this question with many grains of salt, because it's mostly a theoretical curiosity.

I've built a multinomial classifier using GLMnet. The problem is that some of of my input variables have an uneven distribution set values, and the dataset is quite small (around 210 observations). So when I'm building a sparse matrix with these factored variables, the (stochastically partitioned) holdout set will inevitably have fewer columns than the training set, as some factors are not represented. My solution for now has been to make the holdout set quite large (40-50% of all the data), but this seems suboptimal.

I thought one solution to this would be to duplicate every element of the dataset, which would eliminate the problem of non-representation without affecting the relative weights of data points.

However, this didn't seem to work. I still had issues with categorical values not being represented! I will eventually have access to more data for this problem, but I was just wondering a) why this didn't seem to "fool" the sparse matrices, and b) whether this is a permissible thing to do in general.

  • 1
    $\begingroup$ No changing the data this way changes some characteristics of the data and is like a very different sample. $\endgroup$ – Michael R. Chernick Sep 26 '17 at 23:52
  • $\begingroup$ How so? All of the percentages of every quantity are the same. $\endgroup$ – data princess Sep 27 '17 at 1:22
  • $\begingroup$ Larger sample size means that you have greater precision to estimate parameters. The increased sample size is fake and you haven.t really improved anything. The procedure is misleading and creating artificial data should never be done! $\endgroup$ – Michael R. Chernick Sep 27 '17 at 1:51
  • $\begingroup$ Okay, I'm not trying to fool anybody or fake any results. This is so the model has an easier time swallowing it. I'm just looking for a mathematical reason that makes it different. $\endgroup$ – data princess Sep 27 '17 at 2:13
  • $\begingroup$ How is that not what your doing if you are using fake data to fit a model? $\endgroup$ – Michael R. Chernick Sep 27 '17 at 2:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.