@inproceedings{thiesson2005efficient, author = {Thiesson, Bo and Meek, Chris}, title = {Efficient gradient computation for conditional Gaussian models}, booktitle = {Proceedings of Tenth International Workshop on Artificial Intelligence and Statistics}, year = {2005}, month = {January}, abstract = {We introduce Recursive Exponential Mixed Models (REMMs) and derive the gradient of the parameters for the incomplete-data likelihood. We demonstrate how one can use probabilistic inference in Conditional Gaussian (CG) graphical models, a special case of REMMs, to compute the gradient for a CG model. We also demonstrate that this approach can yield simple and effective algorithms for computing the gradient for models with tied parameters and illustrate this approach on stochastic ARMA models.}, publisher = {The Society for Artificial Intelligence and Statistics}, url = {https://www.microsoft.com/en-us/research/publication/efficient-gradient-computation-for-conditional-gaussian-models/}, edition = {Proceedings of Tenth International Workshop on Artificial Intelligence and Statistics}, }