Of an arbitrary number of dimensions
, specified by n. See the warning given on hypercube
Thoughts and models that invole large numbers of dimensions tax the minds of even the greatest thinkers. Computers, however, have no problem. Here is a quote from an introduction to neural nets:
To find the value for each pixel, the average value of the image in the pixel area is found and then thresholded to determine whether it is white or black. We now make the correspondence white = `1', say and black = `0'. This array of Boolean quantities may now be stored in a special purpose computer memory or framestore. Typically the pixel grid may be 512 by 512 giving over 1/4 million pixels. Thus, the pattern space will have dimension 1/4 million. This is often reduced to make things more manageable.
See also: n-dimensional quake