On the Differentiation of a Log-Likelihood Function Using Matrix Calculus
Darrell A. Turkington · UWA Profiles and Research Repository (University of Western Australia) · 2011
Simple theorems based on a mathematical property of vecY/vecX provide powerful tools for obtaining matrix calculus results. By way of illustration, new results are obtained for matrix derivatives involving vecA, vechA, v(A) and vecX where X is a symmetric matrix. The analysis explains exactly how a log-likelihood function should be differentiated using matrix calculus.