I have used the sum of squares differences for a crude comparison of models. Do you think this is a fair method?
What cutoff percentage accuracy would you use for your modeling?
When, how, and why?
There are several methods to measure the performance of a model. Investigate which you prefer.
Which modeling technique appears to fit your data?
Determine a better way to determine whether a binary data model is accurate against a simple percentage correct.
What kind of data would be more suitable to developing a neural net?
Work with the method argument
predictto use other modeling techniques for your data.