It is very useful to upoload the code of yor experiments and let others convince themselves, it is named "reproducible results" and will rise the posibilities of your paper to be accepted, even when the results are not well explained or other typical weaknesses. Best
Sent from my R2D2 El 21/12/2012, a las 15:31, Ben Goertzel <[email protected]> escribió: > > Aaron, > > Anyone can write a paper and submit it to a scientific journal or conference. > There is surely some bias toward submissions from folks with university or > corporate research lab affiliation, but it's not a *huge* bias.... (And > there are other biases too, like the one in most Western journals in favor of > Westerners and against Chinese....) The main issue for independent > scientists without an academic background, is knowing how to write a paper > with the right format and style and the right sort of contents... > > To publish a paper on your new/ improved algorithm, the easiest route is to > show that it outperforms the best current algorithms on some recognized test > problems. If that's not appropriate for your case, then alternatives include > > -- show results of systematic testing of your algorithm on some datasets, > exploring where it works best and trying to explain why, and comparing its > performance to other approaches > > -- prove some theorems about the algorithm.... This looks like "meat" and > then less burden is placed on your experimental results > > Some journals are easier to get accepted in than others. Generally the ones > with higher impact factor (you can find journals sorted by impact factor > various places online) are harder to get papers accepted in. Some > conferences are easier to get accepted in than others too; I think there are > also online ratings of the impact of conference proceedings... But > conference publication has the aspect that you need to go to the conference, > and pay to register, travel, etc.; so it's not really worthwhile unless the > conference actually interests you ... (of course for academics, their > university pays all that...) > > Be aware that most referee reports for journal/conference papers are harsh > and haughty in tone. Don't take it personally, that's just the personality > aspect that refereeing seems to bring out in most academics. Just try to > incorporate whatever constructive criticism is in the referee reports, then > revise and resubmit elsewhere... > > If you would like comments on a draft paper on one of your algorithms, feel > free to send it to me sometime and I'll give you some feedback.... > > thx > ben g > > > > On Thu, Dec 20, 2012 at 12:54 PM, Aaron Hosford <[email protected]> wrote: >> Putting aside any complaints about how the system works vs. how it ought to: >> I've invented a few learning/AI algorithms, and some variants of existing >> ones, too, among other things. As an "independent researcher" myself, I'm a >> little clueless as to how to go about getting published. PM suggested >> joining IEEE or AISB as a means to do so. Any other advice? Is this the only >> route? Do these organizations provide assistance for total newbs? >> >> I aced all my undergraduate CS classes, but lost my scholarship & dropped >> out due to personal issues (homelessness, among other things) before I could >> finish up the non-major classes required to graduate from a state-run >> school, so I don't even have a formal education according to the system, >> despite my rather extreme autodidacticism. Combine this lack of credentials >> with a complete lack of experience and virtually no one to ask for advice >> (you guys are it for me), and I'm not sure how to open this door. Any help, >> insight, or advice would be greatly appreciated. >> >> >> >> Logan wrote: >>> Maybe one day, there will be Open Science, which allows the participation >>> of any being that has knowledge to contribute, (even if they didn't spend >>> tens of thousands and years of their lives apprenticing.) along with Open >>> Studies, where people can participate in a study by doing an experiment at >>> home and posting results, and Open Journals, that actually share the >>> information that science has accumulated. >> >> I have a couple of friends I'm working (very slowly) with to make this a >> possibility. Right now we're just a reserved domain (http://scipubs.com/), >> but eventually we hope to provide a fully functional open access publication >> platform, with a system that literally allows anyone to publish. (There >> will, of course, be certain fundamental standards for publication of a >> particular paper, primarily regarding the style.) Journals, similar to >> channels on YouTube, will review papers, and access to the list of papers >> reviewed and accepted by a particular journal will be visible provided you >> have either paid that journal for access (we take a cut to pay for operation >> costs) or the journal makes their list available for free. (Journals may >> also choose to have authors pay for the review process instead of or in >> addition to readers.) So, in summary: >> >> Anyone can publish for free. >> Anyone can read any article for free. >> The peer review process is still an effective filter for separating the >> serious research from that of cranks and crooks, but this is a value-added >> paid service. >> >> Our hope is that with the ability to support both open access/publication >> alongside the peer review process in the same platform, those who are >> economically, educationally, or opportunistically disadvantaged can still >> contribute to the dialog of scientific progress. Additionally, for >> researchers who are part of the "in" crowd but who are pressured into >> publishing only their most mainstream research, a non-reviewed paper could >> provide an alternative avenue for sharing exploratory/speculative papers >> without having to measure up to the same standards as for primary research >> publications. >> >> >> >> >> On Thu, Dec 20, 2012 at 10:34 AM, Piaget Modeler <[email protected]> >> wrote: >>> Todor, thanks for Stoychev's paper on Developmental Robotics. It's pretty >>> good. >>> >>> The answer is quite simple: >>> >>> Build a system and write an e-book. (Not necessarily in that order). >>> >>> Look at Mentifex (Arthur Murray). He's done it. No bellyaching from him. >>> No complaints. He just works on his system, >>> announces his new developments, and writes his book. I'm very impressed by >>> his commitment. He's a model for us all. >>> >>> If you want to attend conferences and publish papers, join an organization >>> like IEEE or AISB and submit some papers to >>> a conference. Very simple. >>> >>> Life moves on, and because you didn't get credit for some past ideas, >>> that's life. You have to let that go and make room >>> for new ideas. Ensure that you don't repeat those mistakes, and ensure >>> that you DO get credit for your original ideas >>> in the future. >>> >>> The only one stopping you from being all that you can be is yourself. >>> >>> Be great. And let everyone know just how great you are. >>> >>> Cheers, >>> >>> ~PM. >>> >>> >>> >>> AGI | Archives | Modify Your Subscription >> >> AGI | Archives | Modify Your Subscription > > > > -- > Ben Goertzel, PhD > http://goertzel.org > > "My humanity is a constant self-overcoming" -- Friedrich Nietzsche > AGI | Archives | Modify Your Subscription ------------------------------------------- AGI Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-f452e424 Modify Your Subscription: https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-58d57657 Powered by Listbox: http://www.listbox.com
