4 ms·
> Well, a method invented 24 years ago is not exactly novel. Given most things in computer science are older than 1970, I think 1992 counts as being quite new
by bad_user 10y ago
> Well, a method invented 24 years ago is not exactly novel.
Given most things in computer science are older than 1970, I think 1992 counts as being quite new :-)
- merraksh 10y agoSure, but in combinatorial optimization many heuristics appeared after 1992 (variable neighborhood search, feasibility pump, RINS to name just a few).
- geezerjay 10y ago> Given most things in computer science are older than 1970, I think 1992 counts as being quite new :-) Anyone whose remotely interested in heuristics will tell you that heuristics like ant colony optimization are a dime a dozen, and dozens of new heuristics are being published in top scientific journals every single year. Dozens and dozens. Some journal editions even include multiple articles from the same author presenting new heuristics. Things are so out of control in the field of stochastic search methods that this sad state of affairs even originated a theorem that is used to try to justify this loss of focus: the no free lunch theorem. This field is plagued by people who succeed at marketing their own pet heuristic method instead of favouring actual results.