Common-used Formula
Entropy formula of multivariate normal distribution Assume \(x \sim N(\mu, \Sigma)\) and \(x\in R^p\), then the entropy is \[E_x(-\ln P(x; \mu, \Sigma))=\frac{1}{2}\ln|\Sigma| + \frac{p}{2}(1+\ln(2\pi)).\] For more details, please see https://math.stackexchange.com/questions/2029707/entropy-of-the-multivariate-gaussian . Decomposition of Matrix Cholskey decomposition, QR decomposition, LU decomposition, SVD decomposition and Jordan decomposition: see https://blog.csdn.net/mucai1/article/details/85242098 Matrix inverse trick: Assume \(A \in R^{p\times p}\) is invertible, we can speed up the computation by Cholskey decomposition, \[A = L L^T.\]