Hi. Super Mentor @TMosh. I have completed this deep learning specialization. I want to continue learning in this field and indulge in research activities in deep learning .One thing I want to know is how I can keep myself updated in DL, for instance, there are plenty of research papers published each day in journals or conferences. How should I decide which paper to read and know which paper has a high impact or which paper can be a major breakthrough or presenting a novel research idea. As an undergraduate and a novice it’s quite challenging to filter through research papers. Many of the times the research papers are not useful or redundant , if some one reads papers like that its just a waste of time .
1 Like
I don’t have any special guidance on how to decide if a research paper is useful.
Sometimes, a paper that seems unimportant when published will become very significant over time.
If you want to be a DL researcher, you’ll need to be aware of all the current research. That means studying a lot of papers.
Hi, @SHAHID_KAMAL!
@TMosh is right. Research requires a lot of dedication to keep the pace of the state of the art. Nevertheless, if you can have access to the most relevant journals and conference proceedings through your institution (university, work, etc) you can filter by impact factor (JCR, JCI, …).
In terms of accessibiity, arxiv.org does an incredible job at offering almost every paper for free.
Yes I use arxiv.org to access the papers . but what’s the way of knowing the impact factor . can you elaborate on this ?
I think Tom covered this in his earlier response: there is no magic formula. Maybe the only “rule of thumb” is that you will soon have an idea of who the really important researchers are. Any paper that has one of those names on it would go to the front of your list to check. But you just have to dedicate some time to reading all the Abstracts. The good news is that you should be able to tell with a couple of minutes of looking at the Abstract, whether it’s worth your time to read further. That’s kind of the point of Abstracts. 
Sure!
You can check Journal Citation Reports to search for journals and compare their “relative impact on their field”. Journals are divided into four quartiles depending on their relevance (Q1 being the best). For instance, if you search for International Journal of Computer Vision you will see it is one of the best for Artificial Intelligence with several metrics.
Thank you . I got it now