Talk:Combinatorial optimization

Untitled

"Combinatorial optimization algorithms are often implemented in an efficient imperative programming language, in an expressive declarative programming language such as Prolog, or some compromise, perhaps a functional programming language such as Haskell, or a multi-paradigm language such as LISP." - What it actually says: "Combinatorial optimization algorithms are usually implemented in some kind of programming language." -What's the point?

constantin — Preceding unsigned comment added by 84.163.85.41 (talk) 00:30, 20 January 2006 (UTC)

I found the listing of programs and their classifications helpful and useful.

mrxe — Preceding unsigned comment added by 24.85.239.254 (talk) 03:27, 20 April 2006 (UTC)

agree with constantin; that paragraph says nothing

tom — Preceding unsigned comment added by 137.222.102.53 (talk) 15:51, 16 October 2006 (UTC)

I also agree with constantin, and I think that information about implementation would be better located towards the end of the article, as seems to be the style for most algorithm articles in Wikipedia. Davidcoffin 09:44, 18 January 2007 (UTC)

Metropolis and Hill-Climbing Cross References Needed

Besides mention of simulating annealing, the Metropolis algorithm (to be found in Wikipedia, not by "Metropolis" but by "Metropolis-Hastings") should very definitely be referenced or cross-referenced here. More generally, Hill-Climbing should be referenced. And in Hill-Climbing entry, I don't see "probabilistic hill-climbing".
Probabilistic hill-climbing is a counter-intuitive notion. Well, after a while (like when everybody see a unicorn -- so what), the notion seems less counter-intuitive, but in each step you have some chance (probability) of going towards a "worse" configuration -- and in that way you (most likely) avoid getting stuck in local optimum as opposed to global optimum. This central notion should be part of combinatorial optimization entry.199.196.144.11 19:34, 21 March 2007 (UTC)

Add: Agree. This is an important aspect of combinatorial optimization as it is how most real-world applications solve it. In the "Distributed Combinatorial Section," a probabilistic hill-climbing algorithm is cited. — Preceding unsigned comment added by Whatfoxsays (talk • contribs) 21:16, 4 October 2014 (UTC)

Difference between Combinatorial Optimization and Operational Research

Hi wiki community,

I've always wondered what is the difference between Combinatorial Optimization and Operational Research, could someone please bring me some explanation?

Thanks in advance Amine.ensibs (talk) 21:57, 2 October 2017 (UTC)

Operations research is a field of applied mathematics, that is concerned with applications to real-world, usually industrial or logistical problems; it often uses continuous data, thus real values. Combinatorial optimization is arguably foremost a field of pure mathematics, though with many applications, which deals with combinatorial data; this almost always means discrete, for instance finite graphs, finite sets, integers. There is a significant overlap between both fields and sometimes it is hard to decide in which best to classify a given problem - like some types of TSP or some integer programming problems. Plm203 (talk) 09:25, 10 September 2024 (UTC)

This page is probably wrongly classified as part of the series on network science

Although network science treats some combinatorial optimization problems the subject of combinatorial optimization is not a part of network science, and not even tightly linked to network science. Of course in each complexity class we can find hard problems dealing with some types of graph, thus interpretable as network problems, but we can as well find hard integer programming problems, or hard problems on other combinatorial structures. Arguably there should be a series on combinatorial optimization, or one on combinatorial algorithms containing optimization as subseries/section. Perhaps the page on Optimal network design should be used as a replacement for the general combinatorial optimization page, this page, in the network science series. Plm203 (talk) 09:12, 10 September 2024 (UTC)

@David Eppstein: : Your opinion? --Altenmann >talk 16:17, 10 September 2024 (UTC)

Agree. The big distracting navbox is almost entirely off-topic and should be removed. —David Eppstein (talk) 22:38, 10 September 2024 (UTC)