"It's a needle in a haystack problem: 99.99% of the data is noise,"
22 Feb 2005 22:33:32 UTC
Topic 187862
(moderation:
If a real "discovery" is made ie a validated Positive result who gets the Scientific credit? Whats the process? Will all three PC's that crunched the data get mentioned And will the participants get a mention in the discovery.>?
"It's a needle in a haystack problem: 99.99% of the data is nois
)
Somehow I suspect you misread what this project is about. But hey, I won't say anything. :)
> Somehow I suspect you
)
> Somehow I suspect you misread what this project is about. But hey, I won't say
> anything. :)
>
From "Their" Press Release
LIGO's best hope for detecting gravity waves is to spot a cosmic source that sends out regular ripples of gravitational energy. A source such as a spinning star made of neutrons would set the detectors ringing like a bell.
The problem is that the detectors pick up an enormous number of unwanted vibrations. "It's a needle in a haystack problem: 99.99% of the data is noise," says Bruce Allen, a physicist at the University of Wisconsin at Milwaukee.
"This kind of search is not anywhere near possible with our LIGO computing facilities," adds Barry Barrish, head of the collaboration and a physicist at the California Institute of Technology, Pasadena. The data must be analysed at many frequencies, increasing the computer power needed.
So the group is enlisting the public's help. Starting in February, anyone can download a program that will automatically analyse a small chunk of the group's data on his or her personal computer.
If I have "miss read" this and you think its just about crunching numbers I'll leave the project.
I would just like the project to "STATE" what happens.
Sean, I honestly don't think
)
Sean, I honestly don't think the program has gotten that far. We're still working out the finer details of server demand, workloads, lost work units and credits.
A good example of what may happen is SETI, one of the oldest distributed computing projects I know of. If a signal candidate makes it through 'stage 1' then it goes through 'stage 2' and so on, until it is confirmed by the scientific community as legitimate.
I suppose your question is: 'what if you are part of the computing power behind a legitimate signal candidate?' The answer to that is only conjecture, but I think once your computer submits the data back to the lab, it's theirs and you are removed from all credit or recognition, albiet belonging to the greater Einstein (or SETI or Protein) Community. I seems unlikely that any one provider of distributed computing time will get acknowledged at least immediately. Perhaps, after the celebrating and congradulating, someone may dig through the volumes of database information and find out that it was you or your team that actually crunched the data that became a released signal candidate.
For now you are a link in the chain. Hopefully your link is closer to the end of the chain so that we may all see what the end of the chain looks like.
LEX
I know in the Original SETI
)
I know in the Original SETI project, they stated that if you crunched the WU that had the discovery, your name would be placed in the list of co-founders. I would assume this would hold true since SETI has gone BOINC. This does not speak for the other projects, however.
Lex may still have a point. Once the data is crunched and returned, it may very well become their "property" and we may very well relinquish all rights to the data.
Jim
> Sean, I honestly don't think the program has gotten that far. We're still
> working out the finer details of server demand, workloads, lost work units and
> credits.
>
> A good example of what may happen is SETI, one of the oldest distributed
> computing projects I know of. If a signal candidate makes it through 'stage
> 1' then it goes through 'stage 2' and so on, until it is confirmed by the
> scientific community as legitimate.
>
> I suppose your question is: 'what if you are part of the computing power
> behind a legitimate signal candidate?' The answer to that is only conjecture,
> but I think once your computer submits the data back to the lab, it's theirs
> and you are removed from all credit or recognition, albiet belonging to the
> greater Einstein (or SETI or Protein) Community. I seems unlikely that any
> one provider of distributed computing time will get acknowledged at least
> immediately. Perhaps, after the celebrating and congradulating, someone may
> dig through the volumes of database information and find out that it was you
> or your team that actually crunched the data that became a released signal
> candidate.
>
> For now you are a link in the chain. Hopefully your link is closer to the end
> of the chain so that we may all see what the end of the chain looks like.
>
> LEX
>
Jim
The primary investigator will
)
The primary investigator will write a paper and will get a crap load of grant money the next time out. People who designed the system will get co-authors or mentions in the paper. The rest of us will probably get a generic "Thanks to all who helped."
BOINC is pretty much built on altruistic psychology. You're donating unused computer power out of a wish to have the problem solved or a general willingness to help.
This isn't going to turn into any profit making discovery even if we hit gold in two months. So there are no royalties to fight over.
So best case, they find a pulsar, get deluged with grants and use .5% to buy us all a T-Shirt or somthing. That and you actually went out of your way to help the overall cause of discovery. Any further incentive payment would be antithecal to the entire force multiplying concept that makes BOINC work. If they're going to pay for computing power they may as well go rent a CRAY.
As far as i recall, none of
)
As far as i recall, none of us crunchers in this thread have asked for money in the event of a successful find. One might make an argument that we are asking for some recognition (ie our names listed as one of the co-founders in this volunteer project). And to (probably mis)quote Dominique from another thread, "That and $2.00 will get me a nice drink at StarBucks."
Jim
> The primary investigator will write a paper and will get a crap load of grant
> money the next time out. People who designed the system will get co-authors or
> mentions in the paper. The rest of us will probably get a generic "Thanks to
> all who helped."
>
> BOINC is pretty much built on altruistic psychology. You're donating unused
> computer power out of a wish to have the problem solved or a general
> willingness to help.
>
> This isn't going to turn into any profit making discovery even if we hit gold
> in two months. So there are no royalties to fight over.
>
> So best case, they find a pulsar, get deluged with grants and use .5% to buy
> us all a T-Shirt or somthing. That and you actually went out of your way to
> help the overall cause of discovery. Any further incentive payment would be
> antithecal to the entire force multiplying concept that makes BOINC work. If
> they're going to pay for computing power they may as well go rent a CRAY.
>
Jim
> The primary investigator
)
> The primary investigator will write a paper and will get a crap load of grant
> money the next time out. People who designed the system will get co-authors or
> mentions in the paper. The rest of us will probably get a generic "Thanks to
> all who helped."
>
I would be happy just with a list of those that found something somewhere on the Einstein@home site.
If my WU found a pulsar, I would love just to know where it is, what its period is.
After all finding a pulsar or star is not to special, it is what you do with the information. I never could do what "they" do with the equations and analysis. I would not expect my name on a paper, just for running "their" program. But just mention on a list somewhere like those that have high credit number get and information about what my WU found.
If I have the information about what I found and where it is in the sky, I can make my own T shirt.
Dennis
> Why the hell do you feel
)
> Why the hell do you feel like you are entitled to the "discovery"(or even
> mentioning) if you just HAPPEN to analyze the data that sparks discovery?
Why do people feel a need to keep tabs on credit scores?
Recognition of the users whose WU yielded a spectacular discovery would excite everyone involved with the project, and more than likely swell E&H's ranks. I'm not saying we should count on fame and fortune, or even expect a breakthrough during the first few runs (consider Seti@Home), but there are many dreamers out there who like the idea of being recognized.
Why do people sink money into Powerball when the odds are so low? It's human psychology, my friend. When Seti@Home was young, I was excited by the possibility of being recognized for the winning WU myself.
Why do I feel like I am
)
Why do I feel like I am entitled? Because I am donating electricity to this project. I am donating time to download the program and keep it running and updated. I am donating hard-drive space to hold data.
I am not looking for fame or fortune because of a discovery I might or might not make. However, I do feel like mine name should be listed somewhere. Even if it is in the abstract or the footnote of the technical paper.
It could look something like this:
Footnote:
The data from this project was computed using spare CPU cycles of volunteers all over the world. The data that lead to this particular discovery was calculated by Jim Baize, Roy G. Biv, and John Q. Public. The data was then verified by Jane Doe and Mary Palmer.
What happens if / when I don't get that recognition? Well... Nothing. I know I did a good thing. If the author of the paper did not list the volunteers in an effort to minimize the effect of our work on his project thereby making him look like a bigger player in the grand scheme of things, then he will have to live with that on his conscience.
Jim
> > Why the hell do you feel like you are entitled to the "discovery"(or
> even
> > mentioning) if you just HAPPEN to analyze the data that sparks
> discovery?
>
>
Jim
> > > Why the hell do you
)
> > > Why the hell do you feel like you are entitled to the
> "discovery"(or
> > even
> > > mentioning) if you just HAPPEN to analyze the data that sparks
> > discovery?
> >
> >
>
> Why do I feel like I am entitled? Because I am donating electricity to this
> project. I am donating time to download the program and keep it running and
> updated. I am donating hard-drive space to hold data.
>
> I am not looking for fame or fortune because of a discovery I might or might
> not make. However, I do feel like mine name should be listed somewhere. Even
> if it is in the abstract or the footnote of the technical paper.
>
> It could look something like this:
>
> Footnote:
> The data from this project was computed using spare CPU cycles of volunteers
> all over the world. The data that lead to this particular discovery was
> calculated by Jim Baize, Roy G. Biv, and John Q. Public. The data was then
> verified by Jane Doe and Mary Palmer.
>
>
> What happens if / when I don't get that recognition? Well... Nothing. I know
> I did a good thing. If the author of the paper did not list the volunteers in
> an effort to minimize the effect of our work on his project thereby making him
> look like a bigger player in the grand scheme of things, then he will have to
> live with that on his conscience.
>
> Jim
>
Well put Jim, surely it would only be "common courtesy" to recognise the project, all the volunteers involved and particularly those that crunched the final data, no great hardships that, is it ?
Willie