I am currently trying to do a literature review/research to determine if and how well ANNs work on data sets where the number of inputs is smaller than the number of outputs.

For instance, a class project I am thinking of involves a data set with 2 predictors, 18 response variables, and 300 observations. The goal is not to provide an ANN which can "solve" this data set but to properly explore the uses of ANNs against it.

But data sets like these are odd, in my short time as a Computer/Data Scientist I don't think I have seen a data set or problem with this characteristic, let alone a ML/AI method applied to it.

I am looking specifically for research articles on the subject of applying ANNs to data sets with the described characteristic or any knowledge the community might have that relates to this issue.

If not that even a few good keywords to search on, I cannot seem to find anything related to this. Once you start searching for "Neural Network" and "Size" or "Input" or "Output" it all relates to the size of a neural network and what types of problems they can solve, it never describes the data sets themselves.


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.