e_lehman@... a écrit :
> > also have you heard of the "spine" or "backbone" in SAT
>
I may not have understood everything about this notion,
but in two words, the backbone of a formula is the
set of literals implied by the formula.
Some people are studying problems called 2+p-SAT,
p being the probability to generate a clause of size
3, 1-p the probability to generate a clause of
size 2. This idea is to study the transition from
2-SAT in P to 3-SAT in NP.
The notion of backbone has been introduced to
in this framework and has been used later to explain
the behavior of local search algorithms
on random 3-SAT instances.
Some instances with fixed backbone can be found
at SATLIB. They are trivial for complete algorithms
but difficult for local search.
hey TM that is a really beautiful summary/survey, it really gets my neurons going, thanks so much. here's something for everyone elses neurons.. could the...
vznuri@...
Jun 22, 2001 6:49 pm
Hi guys, ... I may not have understood everything about this notion, but in two words, the backbone of a formula is the set of literals implied by the formula....
Daniel Le Berre
daniel@...
Feb 9, 2001 9:21 pm
... DPLL stands for Davis, Logeman and Loveland, aka the paper published in 1962 replacing the directed resolution found in the original paper published by...
Daniel Le Berre
daniel@...
Feb 9, 2001 9:30 pm
... I started using the try-all-variables technique in the late 1980's (88 or 89), at Stanford. It gave some promising results. Alas, I have been unable to...
Dan Pehoushek
danpehous@...
Feb 10, 2001 2:45 pm
Addendum regarding enhancements. Beyond connected components, and try-all-variables at the choice of split-var, I said I was unable to find any enhancements....
Dan Pehoushek
danpehous@...
Feb 10, 2001 3:12 pm
DP-- I was reading a recent survey paper on logic theory, including things like "1st order logic" etc. .. the author had some amazing relations between e.g....
vznuri@...
Feb 11, 2001 12:45 am
... Iso and homo morphisms between different forms of mathematical machinery should not be too unexpected anymore. ... Yes, #P=#Q does have analogs and...
Dan Pehoushek
danpehous@...
Feb 11, 2001 3:38 pm
... I found references to the following papers (haven't seen them), they might be interesting for you: 80b:68050 68C25 Aspvall, Bengt; Plass, Michael F.;...
Marx Dániel
dmarx@...
Feb 12, 2001 8:23 pm
... Boolean formulas. ... fully quantified Boolean formula that has the quantified part in ... algorithm runs in time $O(n+m)$, where $n$ is the number of ... ...
Tom Morrisette
Thomas.Morrisette@...
Feb 12, 2001 9:22 pm
hi TM did you summarize the approach that shows 2Sat is in P over on T-E once, iirc? do you have a ref? did it involve tarjan's ideas or was that something...
vznuri@...
Feb 13, 2001 7:56 pm
... I'm certain I did, yes. But I don't recall precisely when. ... yes, see the two refs in the note you just replied to. :) They both refer back to the...
Tom Morrisette
Thomas.Morrisette@...
Feb 13, 2001 8:42 pm
hi TM .. thanks to the ref to your 2sat article via online email, I dont see why it doesnt deserve posting again ...
vznuri@...
Feb 14, 2001 12:44 am
... You are welcome ... No problem ... According to my vague recollection, the quantification discussion is a good example to emulate in proofs of Dan's result...
Tom Morrisette
Thomas.Morrisette@...
Feb 14, 2001 7:43 pm
... Boolean formulas. ... a fully quantified Boolean formula that has the quantified part in ... algorithm runs in time $O(n+m)$, where $n$ is the number of ...
danpehous@...
Feb 13, 2001 1:21 am
Hi, Vladimir and Tom! ... http://www.geocities.com/st_busygin/sat01_notes.html and http://www.busygin.dp.ua/npc.html Vladimir is right - the method using...
Stas Busygin
busygin@...
Feb 13, 2001 10:44 pm
Hi Stas. Thanks for forming this list. EL: Can you give a reference for the Schoning stuff. It seems very interesting. --Bob Solovay ... assignment to the n...
solovay@...
Feb 9, 2001 10:45 pm
... I couldn't find Dr. Schoning's home page or anything downloadable, but here is the reference Schoning, U, "A probabilistic algorithm for k-SAT and...
Thomas.Morrisette@...
Feb 10, 2001 12:01 am
On Sat, Feb 10, 2001 at 12:01:03AM -0000, Thomas.Morrisette@... wrote: [···] ... [···] Well, he does have something resembling a home page: ...
Niklas Matthies
matthies@...
Feb 10, 2001 12:46 am
Since everyone has been giving an introduction on their interests, I will follow suit. However, my main interests are currently not in CS/algorithms at all. I ...
Lasse
lrempe@...
Feb 10, 2001 6:01 pm
... I think someone else posted the reference. I've put a short postscript summary that I wrote in another context in the files area. It also discusses the...
e_lehman@...
Feb 10, 2001 1:58 am
... Here is an online article that cites Schoning's result: An Improved Exponential-time Algorithm for k-SAT by Ramamohan Paturi Pavel Pudlak Michael E Saks ...
Thomas.Morrisette@...
Feb 10, 2001 2:23 am
... Evidently Schoning's analysis was loose in the preprint. In the final version, he got (4/3)^n and not (3/2)^n as stated above. In fact, it is easy to...
e_lehman@...
Feb 10, 2001 5:03 am
... Sorry to post so much, but I guess I should mention: this is the "nightmare" paper that I referred to before. If you can digest these 44 pages, I salute...
e_lehman@...
Feb 10, 2001 5:11 am
... Laurent Simon's SAT-Ex provides a comparison of a lot of different SAT solvers on various benchmarks: http://www.lri.fr/~simon/satex/satex.php3 Good idea...
Daniel Le Berre
daniel@...
Feb 11, 2001 9:52 pm
Hello Eric, ... How much? Please. :) Anatoly Plotnikov...
Anatoly D. Plotnikov
aplot@...
Feb 7, 2001 8:26 am
hi EL, you wrote over on theory-edge before you left.. [satisfiable coding lemma] ... was wondering you could expand on that: the two lines of development that...