Recommended papers for deep learning based classification?

I'm working on a project where I have to build a classifier using Deep Neural Networks and I will be dealing with big data. I am familiar with neural networks, yet not an expert. I can obviously find a huge amount of literature online regarding deep learning but mostly will be convnets related. Are there any good research papers for DNN-based classifiers or should I just stack more layers? I want to get some insight about the architecture and training process. P.S: Using Deep Neural Networks is mandatory even if the accuracy is low, but it would be great if it worked out.

2 Answers

If you are interested in very recent, highly advanced works I would highly recommend keeping an eye on Google Brain's research paper releases. They have categories from Machine Intelligence, NLP, Machine Translation, Machine Perception, etc.. Many of these research papers involve Deep NN's so it should satisfy your desires.

https://research.google.com/pubs/papers.html

Maybe reading papers is a bit hard, because they usually don't explain everything. I suggest you watching deep learning specialization by professor Andrew Ng, which is brief and easy to figure out. If you insist on reading papers, this three sequential articles may help you. Choosing neural networks highly depend on your task. For most computer vision tasks convolutional neural nets are used. You can also take a look at here which contains different resources and discussions and many other useful stuff.

• thanks for your reply! I'm already familiar with convnets and the first two links your provided. My problem is the lack of literature for using deep learning outside the image-related scope. I'm targeting research literature to get some tips of how should I build my model since I don't need to start from zero. I'll give the book a look and get back to your answer if I find it useful. Feb 12 '18 at 14:22
• @DieDen if that's your case, I recommend you watching the third course of the specialization which has nice guide lines for deploying deep learning tasks. Feb 12 '18 at 14:25