Re: FN-FORUM: "Network" analysis

date posted 24th January 2005 19:32

On 24 Jan 2005 17:44:40 -0000, Charles Lecklider [EMAIL REMOVED]
wrote:

>Alex Farran wrote:
>> Charles Lecklider writes:
>>=20
>>>Starting with the cheapest heap (read: pitch) (or most expensive, or=20
>>>random, or whatever) allocate the longest period of time you can,=20
>>>starting on the day they want. You'll get 0 or more days. Calculate =
the=20
>>>next starting date, pick the next heap. Lather, rinse, repeat.
>>=20
>> That's an example of an heuristic algorithm, where the heuristic you
>> use is the longest period of time available for a pitch starting on a
>> given date. It's quite an extreme example because it relies entirely
>> on the heuristic, with no backtracking to explore other branches. It
>> won't produce the optimal result, but it is fast.
>
>Yes, but how much worse is it than the complex network-based version?=20
>How much easier is it to debug? Would the network version not be a good=20
>example of "premature optimisation"?

A "sub-optimum" solution that returns a value within a few seconds beats =
a
"perfect" one that just skirts the heat-death of the universe.
--=20
Marc Wilson

Cleopatra Consultants Limited - IT Consultants
2 The Grange, Cricklade Street, Old Town, Swindon SN1 3HG
Tel: (44/0) 845 890-3012 Fax: (44/0) 871 236-1531=20
Skype: cleo-marc
Mail: [EMAIL REMOVED] Web: http://www.cleopatra.co.uk
_________________________________________________________________
Try MailTraq at https://my.mailtraq.com/register.asp?code=3Dcleopatra