User:Sintel/Dual Weil-Euclidean norm: Difference between revisions

Sintel (talk | contribs)
more fixes
Sintel (talk | contribs)
Relation to other metrics: -stray subscript
 
(One intermediate revision by the same user not shown)
Line 41: Line 41:


:<math>
:<math>
\begin{gather}
\left\langle \alpha, \beta \right\rangle^{\ast} = \alpha G^{-1} \beta^{\mathsf T}
\left\langle \alpha, \beta \right\rangle^{\ast} = \alpha G^{-1} \beta^{\mathsf T} \\
</math>
 
:<math>
||\alpha||^{\ast} = \sqrt{\left\langle \alpha,\alpha \right\rangle^{\ast}} = \sqrt{\alpha G^{-1} \alpha^{\mathsf T}}
||\alpha||^{\ast} = \sqrt{\left\langle \alpha,\alpha \right\rangle^{\ast}} = \sqrt{\alpha G^{-1} \alpha^{\mathsf T}}
\end{gather}
</math>
</math>


Line 58: Line 59:
<ref>Miller, K. S. (1981). On the Inverse of the Sum of Matrices. Mathematics Magazine, 54(2), 67–72. https://doi.org/10.2307/2690437</ref>
<ref>Miller, K. S. (1981). On the Inverse of the Sum of Matrices. Mathematics Magazine, 54(2), 67–72. https://doi.org/10.2307/2690437</ref>


: If <math>A</math> and <math>A+B</math> are invertible, and <math>B</math> has rank 1, then let <math>g = \text{tr}(BA^{-1})</math>. Then <math>g \neq -1</math> and
::If <math>A</math> and <math>A+B</math> are invertible, and <math>B</math> has rank 1, then let <math>g = \text{tr}(BA^{-1})</math>. Then <math>g \neq -1</math> and


: <math>(A+B)^{−1}=A^{-1} − \frac{1}{1+g}A^{-1}BA^{-1}</math>
::<math>(A+B)^{−1}=A^{-1} − \frac{1}{1+g}A^{-1}BA^{-1}</math>


Now identifying <math>A = W^2</math> and <math>B = j^{\mathsf T}j</math>. We can see that
Now identifying <math>A = W^2</math> and <math>B = j^{\mathsf T}j</math>. We can see that
Line 73: Line 74:
\end{bmatrix}
\end{bmatrix}
</math>, then
</math>, then
:<math> l = W^{-2}j </math>


:<math>
:<math>
\begin{gather}
l = W^{-2}j \\
G^{-1} = W^{-2} - \frac{1}{1+g} l^{\mathsf T}l
G^{-1} = W^{-2} - \frac{1}{1+g} l^{\mathsf T}l
\end{gather}
</math>
</math>


Line 84: Line 84:


:<math>
:<math>
\begin{gather}
j^{\mathsf T}j \circ W^{-2} = I_n
j^{\mathsf T}j \circ W^{-2} = I_n \\
</math>
 
:<math>
g = \text{tr}(BA^{-1}) = \text{tr}(j^{\mathsf T}j W^{-2}) = n
g = \text{tr}(BA^{-1}) = \text{tr}(j^{\mathsf T}j W^{-2}) = n
\end{gather}
</math>
</math>


Line 118: Line 119:
</math>
</math>


Since the metric is equivalent up to scaling, we multiply by <math>\frac{n}{1+E^2_k}</math> to obtain:
Since the metric is equivalent up to scaling, we multiply by <math>\frac{n}{1+E^2}</math> to obtain:


:<math>
:<math>
Line 135: Line 136:
\hline
\hline
k\cdot j
k\cdot j
\end{bmatrix} \\
\end{bmatrix}
</math>
 
:<math>
G(k) =  X_k^{\mathsf T} X_k = W^2 + k^2j^{\mathsf T}j
G(k) =  X_k^{\mathsf T} X_k = W^2 + k^2j^{\mathsf T}j
</math>
</math>
Line 148: Line 152:


:<math>
:<math>
\begin{gather}  
\begin{aligned}  
nk^2E^2 = 1\\
nk^2E^2 &= 1\\
E = \sqrt{\frac{1}{nk^2}}\\
E &= \sqrt{\frac{1}{nk^2}}\\
k = \sqrt{\frac{1}{nE^2}}
k &= \sqrt{\frac{1}{nE^2}}
\end{gather}  
\end{aligned}  
</math>
</math>
== References ==
== References ==