Differences between revisions 3 and 4
Revision 3 as of 2004-02-18 00:05:51
Size: 646
Editor: yakko
Comment:
Revision 4 as of 2005-06-28 02:43:53
Size: 521
Editor: yakko
Comment:
Deletions are marked like this. Additions are marked like this.
Line 5: Line 5:
{{{
                        _____________
                       / n
Geometric Mean = n / __
                     / || Normalized(Pi)
                   \/ i=1
[[latex2($$Geometric~Mean = \sqrt[n]{\prod_{i=1}^{n} Normalized(P_i)}$$)]]
Line 12: Line 7:
     where Normalized(Pi) = Pi / Pn where n is the referenced execution time. Obviously the
     
geometic mean of the referenced machine is 1.
where [[latex2($$Normalized(P_i) = \frac{P_i}{Pn}$$)]] where n is the referenced execution time. Obviously the geometic mean of the referenced machine is 1.
Line 15: Line 9:
}}}
Line 19: Line 13:
{{{
                        GeometricMean(Xi)
GeometricMean(Xi/Yi) = ------------------
                        GeometricMean(Yi)
}}}
[[latex2($$GeometricMean\left(\frac{X_i}{Y_i}\right) = \frac{GeometricMean(X_i)}{GeometricMean(Y_i)}$$)]]

Back to ComputerTerms

The geometric mean of P1,..,Pn is

latex2($$Geometric~Mean = \sqrt[n]{\prod_{i=1}^{n} Normalized(P_i)}$$)

where latex2($$Normalized(P_i) = \frac{P_i}{Pn}$$) where n is the referenced execution time. Obviously the geometic mean of the referenced machine is 1.

One of the nice features of Geometric means is that the following property holds:

latex2($$GeometricMean\left(\frac{X_i}{Y_i}\right) = \frac{GeometricMean(X_i)}{GeometricMean(Y_i)}$$)

Back to ComputerTerms

GeometricMean (last edited 2020-01-23 23:16:12 by scot)