3 ms·
>I'm still trying to develop the intuition for why minimizing E_2 implies taking the mean. Maybe geometric motivation helps. Consider two vectors X = [x2,x2,
by nabla9 6y ago
>I'm still trying to develop the intuition for why minimizing E_2 implies taking the mean.
Maybe geometric motivation helps.
Consider two vectors
X = [x2,x2,...xN],
S = [(x1 - s), (x2 - s), ... , (xN - s)]
How is finding the shortest euclidean distance between X and S related to mean?
- CrazyStat 6y agoI don't understand what you're trying to get at here. Surely the shortest euclidian distance between X and S is obtained by setting s=0, which has nothing to do with the mean.
- yiyus 6y agoIn nabla9's example, S would be the "vector discrepancy". If you want to minimize the discrepancy of s and X (NB. s and X, not S and X, I think this was a mistake), you want this vector to be as small as possible. In other words, you want the magnitude of the S vector to be minimal. This magnitude is given by the norm of the vector, sqrt(sum((S_i)^2)), which will reach a minimum when sum((S_i)^2) reaches a minimum.