mersenneforum.org "New" primality test/check
 Register FAQ Search Today's Posts Mark Forums Read

2018-03-19, 06:21   #265
gophne

Feb 2017

A516 Posts

Quote:
 Originally Posted by ATH Considering how many great minds have worked on this for hundreds of years, most recently Professor Terence Tao, and Dr. James Maynard: https://www.youtube.com/watch?v=MXJ-zpJeY3E https://www.youtube.com/watch?v=QKHKD8bRAro you might want to aim for 17/04/2030 or later to have any non infinitesimal chance at it.
The task is daunting indeed....under normal circumstances 2030 would be optimistic as well!!! more like 3020! would make more sense. Prof Tao and Dr. Maynard are exceptional math geniuses/minds who are able to understand/theorize in all things math.

Myself, I am only going to attempt a "proof" for the twin prime conjecture which I think works, not withstanding that it deals in the realms of infinity.

Are you perhaps suggesting that I should choose the 1st of April, instead of the 17 of April! :) I only chose 17 of April for sentimental reasons because it is my daughter's birthday on that day.

2018-03-19, 06:30   #266
gophne

Feb 2017

3·5·11 Posts

Quote:
 Originally Posted by science_man_88 Another thing to keep in mind is the interconnectedness of conjectures. The twin prime conjecture and the goldbach conjecture are linked by goldbach partitions. So you'll have to potentially look out for provably false implications of conjectures against each other. Not that one will necessarily be false because of the other, just a thing to keep in mind.
Hi Scienceman88

About Goldbach conjecture and other things math I am really very clueless...my big problem is that I am a "prime" hobbyist, and that I have always considered prime numbers to be "ordered" rather than "random". A pattern for Primes of course does exist...a forward pattern! as delivered by operation of the Sieve of Eratosthenes. It is only a quest to find the reverse pattern!

Last fiddled with by gophne on 2018-03-19 at 06:34 Reason: Left out part of a sentance.

2018-03-19, 08:15   #267
CRGreathouse

Aug 2006

5,981 Posts

Quote:
 Originally Posted by gophne The task is daunting indeed....under normal circumstances 2030 would be optimistic as well!!! more like 3020! would make more sense.
Daunting to be sure, but I think there is reason for hope. Much progress has been made recently, and a mathematician who has all of the tools in her/his toolbox will be better able to attack the problem.

The Prime Number Theorem showed the ("trivial") estimate that the gap following a prime p is infinitely often at most length log p or so.

Sieve theory was developed to work on problems like the twin prime conjecture, and the Brun sieve (1915) was developed to prove that the sum of the reciprocals of the twin primes converged.

Westzynthius made the first improvement over the trivial bound, with further (slow) progress through the 1930s by Erdős and Rankin.

Linnik developed sieve theory further with the large sieve in 1941. It uses Fourier analysis and analytic number theory to allow larger sieving sets.

The Selberg sieve was also developed around this time, this one being a combinatorial sieve rather than analytic. Careful choice of sieve weights are essential in this method.

Using the large sieve, Barban, Bomberi, and Vinogradov proved a theorem which gives a Riemann Hypothesis-type error bound to primes in arithmetic progressions when the modulus is up to roughly square root size. The Elliott-Halberstam conjecture (EH) says that the modulus can be taken almost up to the size of the bounding variable itself. This kind of information would be useful in a proof of the twin prime conjecture.

Chen's theorem (1966) was the first near-proof of the twin prime conjecture, limited by the parity problem. It proves that there are infinitely many primes p such that p+2 is either prime or semiprime. Ross published a simplification using the Selberg sieve.

Fouvry & Iwaniec (1980) published an improved version of Bomberi and Vinogradov giving some information above the square root barrier, which in principle makes the twin prime problem attackable.

The Friedlander-Iwaniec theorem (1997) was essentially the first work to begin to break down the parity barrier, using a great deal of analytic number theory along with combinatorics and Fourier analysis.

Various subsets of Goldston, Graham, Motohashi, Pintz, and Yıldırım (~2005; two examples) prove many results about small gaps between primes. One example: On EH, there are infinitely many primes p followed by a gap of at most 16. Sieve weights and Selberg diagonalization are tools.

Green & Tao (2006) published probably the last major twin prime-related paper before the Zhang bomb dropped, and it gives an idea of what the state of knowledge was at that time.

Zhang's theorem, at its heart, is an improvement similar to Fouvry & Iwaniec along with the ingredients needed to apply it to the bounded gap problem. It proves that there are infinitely many primes p followed by a gap of at most 70000000. It may be the largest single step we've taken toward the twin prime conjecture.

The polymath project simplified and improved Zhang's proof and reduced the gap to ~2000.

Maynard added some new ingredients to the proof which allowed a further reduction, then followed up with a new paper with the Tao et al. Polymath project to give a bound of 246 (and a bound of 6 on a stronger form of EH).

Ford, Konyagin, Maynard, Pomerance, & Tao (2018) published a paper, improving on other recent ones I'm too tired to survey at the moment, on long gaps between primes. There seems to be a duality between short gaps and long gaps and this team seems to be unraveling the mystery.

So there really is a lot out there, if you have the patience to learn it. I have quite a lot of learning to do myself.

Edit: Here's a very rough chart of the connections between the steps leading us to current progress toward the twin prime conjecture. All mistakes are mine.
Attached Thumbnails

Last fiddled with by CRGreathouse on 2018-03-19 at 20:50

 2018-04-10, 11:59 #268 gophne   Feb 2017 3×5×11 Posts Twin Prime Conjecture Hi CRGreathouse Thank you for all the information in your post about work done in the area of a possible proof of the Twin Prime Conjecture and related gaps between prime numbers. I will try to look up some of these works to see if I am not going down a path which has already been researched (professionally) and has not managed to nail down a proof for the TPC, as not to be a bore. However, I still intend to post an attempted "proof" on the 17th as I have promised. My attempt would be based on the Sieve of Eratosthenes, Euclid's proof of the infinity of prime numbers, and a logical deduction based on the relationships/statements made/derived from the system of the "proof". What I would like to know is if it is possible to cut & paste text and tables from Excel & Word on this platform, as I have worked mostly in Excel. I hope that if my "proof" is not valid, that at least that it would have been a novel approach to the problem. Even at this late stage I am still struggling with my "logical deduction" on which the proof is based, since to me it appears sound, but there is a fogginess to it as one considers the abstraction of the logic at numbers that cannot be checked in normal computer time (using Excel or SAGE!). The "proof" is not very long/extensive! SO would not waste too much of people's brain power!
2018-04-10, 16:35   #269
CRGreathouse

Aug 2006

5,981 Posts

Quote:
 Originally Posted by gophne I hope that if my "proof" is not valid, that at least that it would have been a novel approach to the problem. Even at this late stage I am still struggling with my "logical deduction" on which the proof is based, since to me it appears sound, but there is a fogginess to it as one considers the abstraction of the logic at numbers that cannot be checked in normal computer time (using Excel or SAGE!).
Please make all efforts to be clear. Don't invent terms unless you must, and if you do you need to give clear, concise definitions. It should be evident at each step precisely what you are claiming and why you think it is true.

Here's my rough scale for grading purported proofs of the twin prime conjecture:
1,000,000 points for a correct proof of the twin prime conjecture
100 points for a clear attempted proof with interesting, nontrivial, correct, and possible publishable theorems/lemmas
0 points for a clear attempted proof
-10 points for a vague, hard-to-follow, or similar attempted proof
-11 points for an attempted proof that is unsalvageable or 'not even wrong'
-11 points for the variants of the standard proof

The average score here, unfortunately, is less than -10.

2018-04-10, 22:17   #270
gophne

Feb 2017

2458 Posts

Quote:
 Originally Posted by CRGreathouse Please make all efforts to be clear. Don't invent terms unless you must, and if you do you need to give clear, concise definitions. It should be evident at each step precisely what you are claiming and why you think it is true. Here's my rough scale for grading purported proofs of the twin prime conjecture: 1,000,000 points for a correct proof of the twin prime conjecture 100 points for a clear attempted proof with interesting, nontrivial, correct, and possible publishable theorems/lemmas 0 points for a clear attempted proof -10 points for a vague, hard-to-follow, or similar attempted proof -11 points for an attempted proof that is unsalvageable or 'not even wrong' -11 points for the variants of the standard proof The average score here, unfortunately, is less than -10.
Noted!

2018-04-10, 22:31   #271
science_man_88

"Forget I exist"
Jul 2009
Dumbassville

100000110010002 Posts

Quote:
 Originally Posted by gophne What I would like to know is if it is possible to cut & paste text and tables from Excel & Word on this platform, as I have worked mostly in Excel.
You could cut paste the document into google sheets, and copy the link for that here.

 2018-04-24, 03:18 #272 gophne   Feb 2017 2458 Posts Hurdle Struggling with my "logical deduction" to show that the methodology I use to locate and predict the location of twin primes, would be applicable at values approaching infinity, or more importantly in terms of the requirements of a proof, that twin primes could still be occuring approaching infinity. If I cannot resolve this shortly, then I will post the "imperfect" proof in order for others to perhaps debunk the approach from their vantage point, if anybody might be interested to do so. I shall post on Google Docs as suggested. Apologies for not posting on the date intended, but I shall post what I have done in a few days time, even if I cannot resolve the problem (proof) myself, just for comment as to the worth or non-worth of the attempt/approach.
2018-04-24, 13:16   #273
CRGreathouse

Aug 2006

10111010111012 Posts

Quote:
 Originally Posted by gophne Apologies for not posting on the date intended
Thanks for prioritizing intellectual rigor.

 Similar Threads Thread Thread Starter Forum Replies Last Post preda GpuOwl 2780 2022-08-09 14:36 Chair Zhuang Miscellaneous Math 21 2018-03-26 22:33 wildrabbitt Miscellaneous Math 11 2015-03-06 08:17 ewmayer Math 11 2007-04-23 19:07 James Heinrich Software 2 2005-03-19 21:58

All times are UTC. The time now is 13:44.

Fri Aug 19 13:44:10 UTC 2022 up 1 day, 11:12, 1 user, load averages: 1.15, 1.47, 1.54