this is for holding javascript data
Vadim Kosoy edited subsection_Orthogonality_Theorem_There_is__.tex
about 8 years ago
Commit id: 74bee67b74b1edec7532961d558df7e77aa03a0b
deletions | additions
diff --git a/subsection_Orthogonality_Theorem_There_is__.tex b/subsection_Orthogonality_Theorem_There_is__.tex
index e6e0929..3e41602 100644
--- a/subsection_Orthogonality_Theorem_There_is__.tex
+++ b/subsection_Orthogonality_Theorem_There_is__.tex
...
Fix $\Gamma$ a growth space of type $(2,2)$ and $\mathcal{E}$ and error space of rank 2. Consider $(\mu,f)$ a distributional estimation problem and $P: \Words \xrightarrow{\Gamma} \Rats$ with bounded range. $P$ is called an \emph{$\mathcal{E}^\sharp(\Gamma)$-optimal predictor for $(\mu,f)$} when for any $S: \Words \xrightarrow{\Gamma} \Rats$ with $\R_S \geq \R_P$
$$\E_{\mu^k \times U^{\R_S(k,j)}}[(P^{kj}(x,y) - f(x))S^{kj}(x,y})] \in \mathcal{E}$$
\begin{equation}
\Abs{\E_{\mu^k \times U^{\R_S(k,j)}}[(P^{kj}(x,y) - f(x))S^{kj}(x,y})]} \in \mathcal{E}
\end{equation}