2
$\begingroup$

I have a problem that I have been working on with logistic regression which involves classifying clients based on their income and socio-demographic variables, all of which are categorical into those who have defaulted on their credit and those who haven't.

I have recently had tensorflow recommended as a package to use for neural networks, however on their website https://tensorflow.rstudio.com/tutorial_mnist_beginners.html they discuss tensorflow in terms of image classification.

My concrete question is thus:

2) Can tensorflow be used for any type of classification problem, for example, a binary classification of observations based upon categorical variables?

Cheers, Steve

$\endgroup$
2
  • $\begingroup$ Hi, why is this still on hold? I edited it to make it more specific... $\endgroup$ – steve zissou Aug 25 '17 at 13:12
  • $\begingroup$ Basic answer to your "can tensorflow be used for (...) a binary classification of observations based upon categorical variables" is YES to the extent that "TensorFlow is a (...) library for numerical computation using data flow graphs". If the specfic built-in function you would like didn't exist, you could write it yourself. $\endgroup$ – tagoma Aug 26 '17 at 20:14
3
$\begingroup$

TensorFlow is a general purpose library for numerical computation using data flow graphs. It is primarily used for neural networks but can be used for any mathematical operations on multidimensional data arrays (tensors).

Thus, TensorFlow can be used to estimate binary logistic regression with explanatory categorical variables. An example can be found in the TensorFlow tutorials here.

$\endgroup$
0

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.