Is the Proof for P=NP Legitimate? Exploring the Controversial Claim

[lufbrajames]

Hi

I found a paper that claims to prove P=NP I can't say that I understand it much, so thought I'd ask for peoples opinions on it.

It can be found here:

http://arxiv.org/PS_cache/arxiv/pdf/1005/1005.3010v1.pdf

Thanks

 

[TMFKAN64]

Lemma 4.4 is bogus. You can't separate any real number into two parts, an integer before the decimal point and an integer after the decimal point.

Also, the possibility that a given computation on an arbitrary input does not halt is not even considered.

 

[Jerbearrrrrr]

Arg, what am I going to tell those traveling salesmen now? Got their hopes up for a sec.

 

[statdad]

Just a comment: if this were a valid proof, why wouldn't it be in a real journal?

 

[lufbrajames]

is arvix not credible?

I only go on there because its free. Perhaps I should steer clear?

Jim

 

[ice109]

Just a comment: if this were a valid proof, why wouldn't it be in a real journal?

perelman's proofs are only up on arxiv

 

[rasmhop]

is arvix not credible?

I only go on there because its free. Perhaps I should steer clear?

Jim

arxiv is a great resource, but the fact that a paper made it onto arxiv does not mean that the paper is good or even correct. Plenty of crackpots manage to get their papers on arxiv.

Arxiv does not try to fill the same role as a respectable peer-reviewed journal and you shouldn't consider it as such. In such a journal you have some assurance that the articles are correct, make sense, and are useful, but on arxiv no such guarantee is made.

Most of the great papers on arxiv is in my opinion either long expository work or pre-prints that'll later be accepted to a standard journal (of course I have only seen a very small subset of the arxiv submission, and in a very narrow area so this may not be true in general).

There are plenty of "proofs" of Goldbach, Riemann, P vs. NP, etc. on arxiv and they all seem to be incorrect.

 

[statdad]

The other commenters have summed it up - the "quality control" at Arvix is pretty poor (if I hadn't already had my morning coffee I would say "does not exist" instead of pretty poor). But another take: If this paper were correct, publication in a journal would be a major coup for the editors. I may be too suspicious, and it may be in review, but if the author truly believed it was correct, and knew that the chance it would appear in a journal were high, why would he leak it out at Arvix?

 

[Mentallic]

ACKNOWLEDGMENT
The author would like to thank all the people who give him worm help and encouragement.

I have a problem to solve, I'd sure like some worm help too