I am an undergraduate who needs to submit a thesis for graduation. I am fairly interested in deep learning, and am working on a project that uses deep learning methods extensively (rCNNs to be precise). The caveat here is that I am working alone, with little help from my advisor. There are hardly any experts on deep learning in my university. How do I go about finishing my project. It is specially intimidating to see that most papers in deep learning are published by multiple accomplished scientists with very little involvement of undergraduate level students.

PS: I am really interested in this topic, so please try not to advise me to switch to an easier topic.

  • $\begingroup$ Don't be intimidated by the position of the authors of the papers. The best resources are usually written by people whose goal is to make the topic understandable. The most important thing is to keep at it until you get it and reach out in forums like this when you need help. $\endgroup$ Jun 10, 2015 at 18:42
  • $\begingroup$ Try to pick something manageable. Decide what a minimum viable model would be for the data you are interested in. Go at it hammer and tongs until done. Start writing your paper based on what you learned. If you get to this stage and still have time left, add more sophistication if you want to. Version control is your friend with experimental software. $\endgroup$ Jun 17, 2015 at 16:09
  • $\begingroup$ @ChristopherLouden: When it comes to deep learning, sadly, there hardly any people who have been working to make the topic understandable. $\endgroup$
    – Rohit
    Jun 18, 2015 at 17:08

2 Answers 2


Coursera and EdX have a courses on Neural Networks (which are the basis of Deep Learning). Stanford also has a good series on Deep Learning.

This question also has some good tips.

For me, what is most helpful is to program the algorithms to help me understand them step by step. Python is good for that purpose.


There are many resources online where you can learn more about deep learning. Specially this link has a great collection. I recently forked this repository. You should go through the resources that interest you. Implementing is the best way to learn, where you get to see the results. As @Christopher mentioned, Pyhton is the best to learn for a beginner. Cheers!


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