The ring $R$ is a ring of probability distributions on $n$ random variables created using markovRing. This method creates a ring map $F: S \to R$ from the ring $S$ of probability distributions on $n-1$ random variables, leaving out the $i$th random variable from $R$. This corresponds to the situation where the $i$th random variable is hidden and $S$ is the ring of observed probability distributions.
|
|
|
This method is frequently used when computing the vanishing ideal of a graphical model with hidden variables by computing the kernel of $F$. For more details see the paper ``Algebraic Geometry of Bayesian Networks'' by Garcia, Stillman, and Sturmfels.
|
|
|
|
|
|
|
|
|
The object hiddenMap is a method function.