I want to classify images in a few different groups with a Neural Network algorithm in R. I have tried different things but nothing works. So my question is: what is the best way to read in images so that I can put the data in a Neural Network algorithm?

Check out my other question as well for things that I have tried already: https://stackoverflow.com/questions/40660306/picture-classification-with-neural-network-in-r

  • $\begingroup$ I know this doesn't answer your question but I would strongly suggest branching into a different language like Python to attempt these kind of tasks, neural networks and specifically convolutional neural networks have a very strong support in a language like Python and very limited in R $\endgroup$ – Jan van der Vegt Dec 14 '16 at 13:26

If you want to use R, loading images can be done with imager. General information, tutorials and case studies can be found here.

As for neural networks in R, the basic ones which are available like nnet, RSNNS and deepnet are to probably too slow. I recommend looking into H2O for deeplearning. If you want to go with GPU support you need to look at other options, like tensorflow or mxnet.

Tensorflow in R: You can use tensorflow from R with the tensorflow package. Information on how to install and use it from R can be found here .

MXNET in R: follow the instructions from the mxnet github page.


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.