Beginner ArXiv papers for non academics?
Hey HN could you please point me towards some of simple arxiv papers to get a hang of reading research papers? Any subject/topic is fine. Most of the papers I look into go way over my head. (I have a bachelors in CS)
1 comments
Please edit your title to use the prefix "Ask HN:" when asking questions.
Reading research papers is an inherently diffucult task. The best advice I was given was to start with a topic you are interested in, then read the full paper start to finish about 10 times, lookimg up unfamiliar terms and math notations while you read. With each iteration you'll gain a better understanding of the topic.
Sometimes, something may seem missing, and that's probably because it is - either by mistake or intentionally (i.e. the authors assume the missing content is understood in the field or is explained at length in a reference). That said, you'll probably need to follow the trail of references.
If a specific subject is new to you (e.g. big data analytics), then before digging into state of the art white papers, I would first search for a text book on that subject.
While Text books generally lag the state of the art, they authors are field experts who distill several years worth of research into plain english and help you form a foundation from which you will then be able to digest state of the art white papers much more easily because you already know the sometimes unique "language" (e.g. special notations, terminology and acronyms) of the field.
For me, simply knowing that I wasn't alone when I felt lost reading white papers for the first time was helpful.
While not Arxiv, if you want to dig into some white papers, here is a good list of links to holdings of recent accepted peer-reviewed papers in the CS field:
https://papers.nips.cc/book/advances-in-neural-information-p...
http://icdm2018.org/program/list-of-accepted-papers/
https://icml.cc/Conferences/2019/Schedule?type=Poster
http://cci.drexel.edu/bigdata/bigdata2018/AcceptedPapers.htm...
https://www.kdd.org/kdd2019/accepted-papers
https://www.ijcai19.org/accepted-papers.html
Reading research papers is an inherently diffucult task. The best advice I was given was to start with a topic you are interested in, then read the full paper start to finish about 10 times, lookimg up unfamiliar terms and math notations while you read. With each iteration you'll gain a better understanding of the topic.
Sometimes, something may seem missing, and that's probably because it is - either by mistake or intentionally (i.e. the authors assume the missing content is understood in the field or is explained at length in a reference). That said, you'll probably need to follow the trail of references.
If a specific subject is new to you (e.g. big data analytics), then before digging into state of the art white papers, I would first search for a text book on that subject.
While Text books generally lag the state of the art, they authors are field experts who distill several years worth of research into plain english and help you form a foundation from which you will then be able to digest state of the art white papers much more easily because you already know the sometimes unique "language" (e.g. special notations, terminology and acronyms) of the field.
For me, simply knowing that I wasn't alone when I felt lost reading white papers for the first time was helpful.
While not Arxiv, if you want to dig into some white papers, here is a good list of links to holdings of recent accepted peer-reviewed papers in the CS field:
https://papers.nips.cc/book/advances-in-neural-information-p...
http://icdm2018.org/program/list-of-accepted-papers/
https://icml.cc/Conferences/2019/Schedule?type=Poster
http://cci.drexel.edu/bigdata/bigdata2018/AcceptedPapers.htm...
https://www.kdd.org/kdd2019/accepted-papers
https://www.ijcai19.org/accepted-papers.html