558
个编辑
更改
→高斯噪声迭代系统基于SVD的因果涌现
</blockquote>
</blockquote>
该策略可以让我们得到的宏观态的近似可逆性，比只使用<math>W=U_1</math>或<math>W=V_1</math>更接近真实的基于有效秩删减奇异值后得到的近似可逆性。
该策略可以让我们得到的宏观态的近似可逆性，比只使用<math>W=U_1</math>或<math>W=V_1</math>更接近真实的基于有效秩删减奇异值后得到的近似可逆性。
==高斯噪声迭代系统基于SVD的粗粒化策略进阶版==
<blockquote>
<math>
\tilde{\Sigma}=\begin{pmatrix}
A^T\Sigma^{-1}A & O \\
O & \Sigma^{-1} \\
\end{pmatrix}
</math>
</blockquote>
<blockquote>
<math>
\tilde{\Sigma}=\tilde{U}\tilde{S}\tilde{U}^T
\end{pmatrix}
</math>
</blockquote>
<blockquote>
<math>
\tilde{U}=(\tilde{U}_{1},\tilde{U}_{2}),\tilde{U}_{1}\in\mathcal{R}^{2n\times r_{\epsilon 0}}, \tilde{U}_{2}\in\mathcal{R}^{2n\times (2n-r_{\epsilon 0})}
</math>
</blockquote>
<blockquote>
<math>
\tilde{\tilde{U}}_{1}=\tilde{U}_{11}+\tilde{U}_{12}
</math>
</blockquote>
<blockquote>
<math>
\tilde{\tilde{U}}_{1}\tilde{S}=\hat{U}\hat{S}\hat{U}^T
</math>
</blockquote>
<blockquote>
<math>
\hat{U}=(\hat{U}_{1}, \hat{U}_{2}),\hat{U}_{1}\in\mathcal{R}^{n\times r_{\epsilon}}, \hat{U}_{2}\in\mathcal{R}^{n\times n-r_{\epsilon}}
</math>
</blockquote>
<blockquote>
<math>
W=\hat{U}_{1}^T\in\mathcal{R}^{r_{\epsilon}\times n}
</math>
</blockquote>
==连续系统具体案例==
==连续系统具体案例==