Anyone looked at "Traveling Salesman Problem" Boinc project yet ?

[Da-Grizz]

TSP (Traveling Salesman Problem) seems pretty stable - until KWSN and Seti USA got involved and broke it (ran out of work over Xmas :rolleyes: with their fight for the No one spot in a one week SURGE period until Christmas :rolleyes:)

TSP http://bob.myisland.as/tsp/

Suitable candidate for Vault consideration ?

Da-Grizz (KWSN Regds) :D

[russkris]

Yes I have but really I cant see the point of the project but having said that, that isn't a rule, so my thought dont count...

Maybe you could tells us a bit more....eh?

[mgpower0]

Running my first w/u at the moment, long (around 23hrs) on a 2.2ghz c2d macbook. Their is a plus though (w/u's for all platforms), and really the results are probably more useful to a lot of people than finding a prime number with lots of digits?????:D Basically the project is designed to find the shortest road route between 40 something cities in the US of A

[russkris]

:ninjawel: mgpower0, especially an Aussie

Thanks for you info..

[mgpower0]

Thanks for the welcome, been a member of forum since April, can't believe I haven't posted till now:confused:

[russkris]

What team you crunch for mate....? Possibly BOINC@australia?

No team = Join Team Ninja :D

[mgpower0]

Nah, part of Crunchers Inc

[vaio]

Nah, part of Crunchers Inc

Me too :D

I spose this project is good re climate change.......creating energy efficient travelling salesmen. :D

[umccullough]

The ultimate goal of the project is actually to develop effective "search" algorithms to solve problems of this type. Currently the project runs both a brute-force algorithm as well as a genetic algorithm.

It's clear from the current results that the genetic algorithm is far smarter ;)

click between brute force and genetic at the top left:
http://www.myisland.as/tsp/index.php

As mentioned in their message boards, the project may eventually be renamed to "The Search Problem". The Genetic Algorithm currently being tested is ultimately planned to be open-source apparently.

[Da-Grizz]

Ni and Hi !

Lots of aplications really - not least being printed circuit board routing layout , aircraft and NASA craft cable/hydraulic routing , road vehicle interals and trafic control - many needs for an efficiant path solution that are done by trial and sucess ie a brute force solution .
This project has set out to discover the best path (40 US Cities) by brute force , and then to develop a general algorithym from knowing the best brute force solution and testing them against a known good solution .
The problem is HUGE - think only 40 destinations in the project as against the number of destinations on a PCB , and it still has to be a DC endeavour for the 40 cities !!!

Regds Da-Grizz

[Al Dente]

I wholeheartedly agree with Da-Grizz, a very useful project. It seems pretty stable with problems sorted very quickly, plenty of work (normally), no participation restrictions, so it seems to meet all the Vault criteria.

And of course, the fact that KWSN would gain a whole heap of points has nothing to do with it.:bateyes:

[russkris]

This is great, Thank you everyone... We have gotten so much feedback for project just recently... It make it really each to add project when people help out with the UP's and DOWN's of a project that I am watching to add to the Vault....

Thank you everyone again

[Da-Grizz]

This is great, Thank you everyone... We have gotten so much feedback for project just recently... It make it really each to add project when people help out with the UP's and DOWN's of a project that I am watching to add to the Vault....

Thank you everyone again

You are welcome Sir .

And many thanks for the discussion areas in APS and Valt inclusion rules , the posters know who they are .

Regds KWSN Da-Grizz

[Da-Grizz]

Ni !! :)

TSP has been stable now for over 2 months and all the big DC Vault teams have a presence there .
More users and teams than other DC Vault projects

Worth another look ?

Regds Da-Grizz

[KAMCOBILL]

Looks good from here too.

[cswchan]

So far so good... so far the WUs are averaging about an hour each... no shortage of WUs... so far... ;)

[russkris]

OK. so she is in the vault, sorry about the delay, I stuffed up a file name..:lol: ooooops

[fractal]

from http://bob.myisland.as/tsp/

2007-04-14
When it rains it pours. I'll be taking a month break from TSP while I try to find work, and fix some projects that have fallen appart here. I'm very sorry ot leave you all in a this state, but it's where I amSo, with no work since 26-Mar and none forthcoming, McCoy would say "It's dead Jim"

[Da-Grizz]

Hmmm - Work is still trick'lin in (See Free Dc Stats) . Similar to the BBC Climate Prediction problem perhaps - that is still accepting results after over 6 months of no new work.
Too early to judge yet I would say .

Regds Da-Grizz [Member of KWSN]

PS : If anyone thinks that TSP is My favorite project , I urge them to peruse the stats and see what Projects I contribute to (under my own name of course ;)).

[glennpat]

I have had 10 boxes on it since it was vaulted and did not getting any work during that time. Its not looking good for some time in getting work. We only had a few hours of work on the project before it was vaulted as a result we're taking a big hit in the vault. Wish I would of crunched some more awhile back.