mind/soul

The Walrus the_walrus at BLOCK.bigfoot.com
Fri Oct 30 09:07:15 EST 1998


Rick Harker wrote:

> > Someone wrote:
> > >>If N neurons have to meet up accurately with N locations, there are
> > >>N! (that's a factorial) ways of doing this.  Given that N is very
>
> I've haven't learned about factorials yet.
> But shouldn't this be ((n*n)-n)/2?
> (for the number of connections? It would eliminate counting a path both
> ways....)

No ... that's O(n^2), whereas factorial algorithms are O(x^n), for some
constant x ...

If you eliminate counting both ways, you are essentially dividing by a
constant factor, not changing from an exponential algorithm to a polynomial
...

d.







More information about the Neur-sci mailing list