I'm working in the medical field and I'd like to learn applications of CNN for image recognition and classification. All the (few) things I learned come from self-learning on the web or sparse books. I'm studying now Tensorflow for CNN implementation but I'm having trouble finding clear documentation for my actual level, so I think I'm missing the basic knowledge in order to understand this.

I'm at a basic level of python programming, I have better understanding of classical machine learning algorithms, which resources should I learn in order to get a good grasp of the argument? Is there such an ideal pathway to this?


1 Answer 1


I'm assuming you're talking about medical images classification, rather than localization.

Personally I recommend Kaggle, it has an awesome forum and people share their codes and opinions there.

You can start at Digit Recognizer, it's actually the well-known MNIST dataset(hand-written numbers). There's no relationship between MNIST and medical field. However there are some common techniques and tricks as they are both image recognition/classification problems.

If you encounter problems, read the other ones' codes at the kernel section, especially those with upvotes.

In the meanwhile, comment section is a good place to learn. I learned a lot there when I started to learn CNN.

  • $\begingroup$ Thanks for the suggestions, do you know more structured sources of information like books? I'm having a difficult time learning from spot sources $\endgroup$
    – GGA
    Feb 4, 2017 at 9:53
  • $\begingroup$ For structured sources, I highly recommend Stanford's CS231n. It's an amazing lecture with details. You can checkout the videos at youtube. And Andrew Ng's Machine Learning Yearning might help you when you become familier with neural networks. $\endgroup$
    – Icyblade
    Feb 4, 2017 at 10:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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