I'm cross posting this to the miniKanren mailing list. It sounds like core.logic or cKanren could be applied to your problem but would need more details. I also wouldn't rule out finite domain solvers like JaCoP.
David On Wed, Mar 5, 2014 at 9:52 AM, Laurens Van Houtven <_...@lvh.cc> wrote: > Hi! > > > I've been experimenting solving some real-world problems related to > organizing a sizable (2000-3000 people) programming conference with a > strong open source flavor. > > My next problem is a bit more daunting. > > This conference has a financial aid program. People who can not afford > to come to the conference are given grants that have to come out of a > budget. The budget is much smaller than the sum of all requested > grants, so it turns into an underspecified optimization problem. In > order to make that issue slightly less bad, there's a score function > so that we can bias in favor of certain people, e.g. speakers. > > I'm convinced the optimal solution, even assuming it exists, is > impossible to find. The data is incomplete; it is not necessarily true > that because you assign less than what people have requested that they > won't accept the grant. > > I wrote an implementation that plausibly matches the above: > > https://github.com/lvh/hood/blob/master/src/hood/core.clj > > I'm sure that that code has a lot of flaws. I haven't really played > with it enough yet. Code review on that is very welcome. Any > suggestions for a cleverer allocation algorithm would also be much > appreciated. > > That said, I really sent this e-mail because I was wondering to what > extent core.logic could be applied to this problem. The "running your > program in reverse" concept is potentially interesting here; I'm > hoping that we could get answers to questions like "what would an > application have to look like to get X dollars under our budget?" > > I seriously don't even know where to start. What would the logic > variables be? It seems that a lot of the example code uses a single > logic variable and then "structures" it in the let body; but I have no > idea what that one variable would even mean. > > Perhaps I should just tackle some simpler logic programs first :-) > > Thanks in advance for your thoughts and comments! > lvh > > -- > You received this message because you are subscribed to the Google > Groups "Clojure" group. > To post to this group, send email to clojure@googlegroups.com > Note that posts from new members are moderated - please be patient with > your first post. > To unsubscribe from this group, send email to > clojure+unsubscr...@googlegroups.com > For more options, visit this group at > http://groups.google.com/group/clojure?hl=en > --- > You received this message because you are subscribed to the Google Groups > "Clojure" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to clojure+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/groups/opt_out. > -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en --- You received this message because you are subscribed to the Google Groups "Clojure" group. To unsubscribe from this group and stop receiving emails from it, send an email to clojure+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
