I am a pre-final year student pursuing CSE. I have developed a machine-learning algorithm, and now I really want to get it published in a peer-reviewed journal. (I’m not exactly sure if the algorithm already exists).
Research papers that document a new algorithm (especially in the area of machine learning) usually go about comparing their algorithm with existing ones to show the performance difference and list out the areas in which their relatively new algorithm will be useful in.
Do I have to implement the existing solution to the problem I’m trying to solve for comparing its performance with my algorithm? Or is it possible to shorten this step by using an existing code?
Is it absolutely necessary to compare the performance? [Because there are just too many choices and the difference my algorithm is going to cause is quite obvious].
Note: I have neither published papers before nor have any professional support to guide me, I’m doing all this on my own. So please excuse me if this question is quite basic.
I have a hard time understanding statements such as:
Scientist X discovered Y in 1960 and was subsequently rediscovered by Scientist Z in 1980.
How does one prove that plagiarism has not taken place? This is also extensively observed in some really old math theorems, chemistry and so on. How does one prove that the similar work produced was as a result of one’s own independent work and not resulting from another’s. Even if something was re-discovered subsequently, why is it even given merit?
Note: this question was already asked at MathOverflow but deemed off-topic. I think this is a better place to post it.
The purpose of a research paper is to present new results in a given field. But how far should an author go in developing his results? Since the discovery of a single new theorem could affect a variety of other fields/subfields, should the author go a step further by dedicating one or more sections of the paper to discussing several implications of the paper’s main result?
Or should the paper “draw the line” once the main results have been established, and let other authors derive its implications in further publications?
What does it means a project being licensed with the name of the author?
I mean, one instance is: there is an author X, that writes an open source book on biology with git and at the end of the README file is © 2017 X. Other instance is when the same author X, writes his webpage and at the end of the page put his characteristic © 2017 X.
It is evident author does not get into the game of licenses (justified or not!!). But what should I interpret with his license? Every change (in any sense) that I want to make I should contact him? What is the (legal status) point of this license “© 2017 X“??
Example of webpage
A hypothetical situation: two (groups of) authors have independently proven the same result. How do we determine who gets credit?
There are some obvious bounds – if Group A posts a paper to the arxiv, and Group B posts the same result to the arxiv n time units later and claims to be independent; if n=10 years, then clearly Group A gets credit. If n=1 week then it’s less clear to me.
Other than posting a paper to the arxiv, to what extent do talks count? What about talks containing nearly complete proofs?
ETA: I’m aware that such situations occur not infrequently; I only say “hypothetical” in the sense that I’m not personally involved in such a situation and I’m just curious to know the community’s opinion.
Imagine someone has proven a interesting mathematical result, but notices that this result contradicts existing literature.
Does the author have to find an explicit error or counter-example to the existing literature before considering submission ? Can she submit the proven result to a journal, if she cannot find an error in her own proof?
I’ve had a little corollary sitting around a while now. It is closely related to a major theorem in my area. Its statement is easy to understand (and rather interesting/amusing on the surface of it) and it’s proof can be written in a couple of paragraphs.
I thought it had potential to be the beginnings of a nice paper, but I’ve struggled to find good “related stuff” to add to it.
Should I just submit this little 1.5 page paper, or would I be wasting everyone’s time? Do any good journals regularly accept these kinds of “notes”?
The journal Immunity says on this webpage
“As a matter of publishing ethics, we cannot consider any paper that contains data that have been published or submitted for publication elsewhere.”
Why not? What if the two papers analyze rich sources of data that overlap only partially? What if they have completely different ambitions? Is the only meritorious activity in science the act of gathering data, and to hell with the analysis? And what does this have to do with ethics?
I am an undergraduate student currently working in a co-op position at a company in my field. I was lucky enough to have the opportunity to be the lead author on a paper based on the research I performed here, and I have a couple of questions regarding what I can do with this information.
I would like to publish it while also making it as visible as possible (partly because I consider myself a proponent of open sharing of information, but also because I’d like to refer future potential employers to it and it would be easier if it were not behind a paywall). I understand that some journals have open access options, but these tend to be quite expensive. What options do I have for publishing in an open way without expensive costs (ideally no cost, since my company has made it clear that they will not financially support publication)?
I understand that one solution to this could be putting it on ArXiv. However, since ArXiv is not a journal, would it be considered incorrect or misleading to call my paper a publication if I were to post it there?
My last question is with regards to contact information in papers. Since I will be leaving the company in a few months, I would like to be able to put an email address in the contact information for the paper that will still be active for a long time. Is it frowned upon or does it look bad if I use my personal Gmail account for contact information?
Thank you for all your help!
In most engineering papers, I see that the authors justify their research by displaying how well it performs against the state-of-the-art methods or algorithms, in a quantitative sense. To put it loosely, this is usually stated along the lines of “Our method is faster/better/cheaper than existing methods, and here are the charts and graphs to prove it”.
However, what about papers that are putting forward a more visionary approach that would take time to justify using quantitative means? For example, it is conceivable that there is an idea that performs moderately well, but needs work by the entire community in order to reach its potential. It is also conceivable that this idea is a more fruitful one than existing lines of research.
Does a piece of research have to beat others in a measurable way in order to be publishable, or is there also a qualitative, holistic way that papers are evaluated?