Wikipedia:Reference desk/Archives/Mathematics/2006 July 19

Humanities Science Mathematics Computing/IT Language Miscellaneous Archives

The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions at one of the pages linked to above.

< July 18 Mathematics desk archive July 20 >

primes now

hi/ Haw I can publish my researches in any scientific journal I have got many results as:
(1)I prove the infinity of Mersenne prime numbers and find it as big as we want.
(2)I prove the infinity of Fermat’s prime numbers and find it as big as we want.
(3)I prove the infinity of the twin primes and find it as big as we want.
More than this I have please help...

Salim Ghodeif

You may wish to start by studying English grammar. --Alexs letterbox 09:43, 19 July 2006 (UTC)

I can assure you that if the results hold up under scrutiny (I have to confess I have some doubts there), any mathematical journal will be abundantly happy to accept papers on any of these results and provide for free editorial assistance in improving the English. --Lambiam^Talk 10:27, 19 July 2006 (UTC)

Take a look at the academic publishing and scientific journal articles to get an idea of the process you need to go through when publishing an academic paper. Note that every reputable journal operates a system of peer review, where each paper is scrutinised by experts in the relevant field before it is published. If you don't like the thought of this, then maybe academic publishing is not the right route for you. Note also that each journal will insist that a submitted paper conforms to a set of submission standards before it will even send it out peer review. Getting your first paper up to the required standard can be a lengthy process. In view of the ambitious nature of your claims above, you might want to get a qualified mathematician to informally review your ideas before you go to the trouble of writing a formal paper. Gandalf61 10:09, 19 July 2006 (UTC)

Hi. Dumb question about wiki viewing.

I'm using IE and Firefox 1.5.0.4 . Unfortunately I can't see correctly some pages such as: http://godseye.com/wiki/index.php?title=Chebyshev_polynomials Instead of formulas I see "diamond math" marks -- \<math\>. I tried to get hint on Wiki pages but I failed. Could somebody point me to such page or tell what soft should I install to view similar pages correctly? Strangely enough on this page all expression with math marks are functional in show preview mode.

Thank you in advance.

I'm really not sure about that web site. You're not having this same problem here on Wikipedia, are you? —Mets501 (talk) 14:08, 19 July 2006 (UTC)

When I look at Chebyshev polynomials on Wikipedia, all is well; when I look at the copy of that page at godseye.com, the appearance is as you describe. Therefore, there is nothing to do on the browser side; that wiki is not configured properly to run texvc so that equations are converted to PNG images. I suggest you communicate with whoever runs that site, not us. --KSmrq^T 14:20, 19 July 2006 (UTC)

Thanks guys. On mark. Looks to be indeed a site problem...... wiki site is ok.

stocks&shares

if a shopkeeper allows me 10% discount and i pay him 21.6rupees,what is the original price. —Preceding unsigned comment added by 59.92.113.39 (talk • contribs) 11:50, 2006 July 19

• Please do your own homework. If you get a 10% discount, then what percentage do you pay for? If you know that you can calculate what the original price was. -- Mgm|^(talk) 13:25, 19 July 2006 (UTC)

By my reckoning, the original price was approximately 1.89 ringgits. With any luck it will take you more time to convert that to rupees (which kind?!) than it would to do your own homework. Please read the instructions for posting at the top of this page, and never again ask us to do your homework for you. --KSmrq^T 14:55, 19 July 2006 (UTC)

Distinguishing two Bernoulli distributions

I have a coin that is either fair (.50-.50) or biased (.51-.49). How many times do I have to flip it before I can tell which it is with, say, 90% confidence? I remember there's a formula for this and there's a wikipedia article about it, but I'm having trouble finding that article from the obvious starting points. The Kolmogorov-Smirnov test and its relatives are too general. No this is not homework, I want to write an article about randomness extraction (see min-entropy and the entropy distillation section of one-time pad). Phr (talk) 20:33, 19 July 2006 (UTC)

In order to solve this, you need to know the apriori probability of having a fair coin. In the simplest case, this would be 50%, but in general it can be anything. Once you know this, there is perhaps some simple formula for working your question out, but that isn't necessary - a straightforward brute force calculation seems to be sufficient (using Bayes' theorem, etc.). -- Meni Rosenfeld (talk) 10:37, 20 July 2006 (UTC)

Here's an approach. If the coin is fair you expect to get .5n heads in n tosses; if it is biased your expectation is .51n heads. So let's say you split the difference and decide the coin is fair if you get less than .505n heads, and biased otherwise. In either case you want a less than 10% chance of making an error. So let's choose n so that the probability of getting .505n heads or more with a fair coin is less than 10%, and the probability of getting .505n heads or fewer with a biased coin is also less than 10%. Approximating each Bernoulli distribution with a normal distribution with variance np(1-p), we find that the variance for the fair coin is .25n and the variance for the biased coin is slightly less. We want 10% or less of the normal distribution to lie above (or below) .505n, so we want .005n to be greater than 1.28 standard deviations. By my calculations, n has to be at least 128²=16384. Gandalf61 13:38, 20 July 2006 (UTC)

I read in my statistics book about some French guy who actually went to the trouble of trying something like this out - he actually flipped a coin several thousand times and counted the results. Unfortunately, he didn't flip it enough times to guarantee it was within such a narrow margin of 50/50. Black Carrot 16:09, 20 July 2006 (UTC)

Need help with extremely slow computer

As of lately, my laptop computer has been acting excruciatingly slow, and I don't know why. I am not sure if it is a memory problem or something more serious. It never used to do this, but now it takes long to do anything, even when I'm not connected to the Internet. Today, it took 7 minutes to connect to MSN, and I have broadband connection with an Ethernet cable! When I right-click on the desktop, it takes several seconds for a menu to pop up.

It is a Compaq Presario 2100 with Windows XP. I have ruled out spyware because I have the Webroot Spy Sweeper program, which takes care of that. When I go to "My Computer" and hover my cursor over the "Local Disk (C:)" icon, it says I have 16.3 GB of free space and 27.9 GB of total space. Do I have enough memory left, or do I need more in order to let my computer run faster? ⇒ JarlaxleArtemis 23:48, 19 July 2006 (UTC)

I would start by defragmenting the disk. This can be done by going into Start->Programs->Accessories->System Tools->Defragment Disk. Then just click "Defragment". If that doesn't work, you can always consider the option of adding more memory. Also, the numbers you mentioned don't have to do with speed or memory. That's merely the amount of information on your computer. Your memory is measured in GHz, not Gb. Whenever a program runs, it often has to store information for short periods of time for its own reference. This is what memory is for. When there isn't enough space for all of a program's "variables" in the memory it will run slower. -- He Who Is^{[ Talk ]} 01:02, 20 July 2006 (UTC)

Is there a risk of losing documents when defragging? ⇒ JarlaxleArtemis 03:03, 20 July 2006 (UTC)

No. And memory is measuring in MB or GB, not GHz. GHz measures your processor speeds (which stays constant, well, since you may have a mobile processor it may vary but this wouldn't account for your problems). Your hard drive space (what you see when you go over your "Local Disk" isn't going to be the bottleknock either. Try right clicking on the taskbar at the bottom of your screen and clicking on Task Manager, then switch the task manager tab over to processess and report how many processess you have running (a count is kept in the bottom left corner). --droptone 03:14, 20 July 2006 (UTC)

It says I have 54 processes running. ⇒ JarlaxleArtemis 03:17, 20 July 2006 (UTC)

That is a few too many processes and is probably overloading your processor (and may be the cause of this recently slowness). I'm sure there is an easier way to do this but I'll give a recommendation. What I'd do is have Task Manager open and type the the name of each process into Google and discover what it actually is. Here is a screenshot from the Task Manager on my computer. So I'd type in AGRSMMSG.exe (or just AGRSMMSG) into Google (or your favorite search engine). The first return I see is this, and it says this process is running because of the 56K modem that is installed on this laptop. Going through this for each process will help you understand what is running in the background of your computer and why. Now in order to stop some of these excessive programs from running you can right click on the process in your task manager and select End Process. In order to stop the process from running in the future you need to open the start menu, select run, then type in msconfig. Now click on the very last tab that says Startup. This displays the list of the programs that start when you start your computer. Some programs place pre-loaders into this list. They are meant to speed up loading times for those programs, but if you hardly use the program then a pre-loader isn't helping you. For example, here is a screenshot of msconfig on my computer. The last item, InterVideo WinCinema Manager, is a preloader for WinDVD which politely installed itself when I installed WinDVD onto this computer. I deselected it in order to stop it from loading when my computer starts up because I do not watch that many DVD's on my computer to justify that process always running and using up my processor's resources. But I also have AIM selected because I do want AIM (AOL instant messenger if you did not know) to load when my computer starts up. This process is long, annoying, and I'd probably err on the side of caution if you don't know what something is. Of course you could mess up your computer by doing this, so I'd only deselect items for programs you use (like WinAMP, WinZip, Adobe Acrobat Reader all have preloaders that are in your system tray and use up your processor) and not programs needed for Windows (for example explorer.exe is the Windows File Manager and ought not have its process ended).--droptone 04:18, 20 July 2006 (UTC)

Additionally, you should defragment your computer in order to have optimal performence (my mom's computer once was insanely slow and after a defragment ran much faster), and reboot every so often (every couple of days if it isn't a problem). These steps are easier than a more drastic measure like reformatting and reinstalling Windows.--droptone 04:22, 20 July 2006 (UTC)

The first step in approaching a problem like this is common sense. If you didn't suddenly change the amount of hard drive capacity or main memory space on your computer, if you didn't swap CPU's, then hardware is probably not the problem. Did this change in behavior come on overnight, or has the slowdown been gradual? If the change was sudden, can you recall what you may have done immediately before? Did you install some new software, such as a screensaver? Such dramatic slowness is a symptom that suggests you are no longer in control of your computer; you may have been compromised. Maybe it's spyware that Spy Sweeper didn't catch; try other free testers. Maybe it's an infection; try free anti-viral software like AVG, and be sure its database is current. Some infections can be deep and insidious, tucked away in a boot sector or as a root kit. Checking the list of tasks is good, but some parasitic CPU-sappers don't show up as tasks.

And not to belabor the obvious, but in many cases the best long-term solution is to switch to a better operating system, one that doesn't constantly invite these hair-pulling exercises. This is the conclusion of Sophos, a major security firm, in their 2006 first half Security Report. --KSmrq^T 04:59, 20 July 2006 (UTC)

Thanks, I defragged it, and it's working a bit faster now. ⇒ JarlaxleArtemis 20:56, 23 July 2006 (UTC)

The computer quickly returned to its slow ways in a couple of days, but I know what the problem is now. My uncle, who is a computer developer, looked at it and figured out that it needs more RAM. ⇒ JarlaxleArtemis 23:43, 6 August 2006 (UTC)