We investigate the algebra and geometry of general interventions in discrete DAG models. To this end, we develop the formalism to study these models as subvarieties of multiprojective space and introduce a theory for modeling soft interventions in the more general family of staged tree models. We then consider the problem of finding their defining equations, and we derive a combinatorial criterion for identifying interventional staged tree models for which the defining ideal is toric. This criterion, when combined with a new characterization of decomposable DAG models in terms of their associated staged trees, specializes to a graphical criterion in the case of discrete interventional DAG models.
翻译:我们调查了离散的DAG模型中一般干预的代数和几何方法。为此目的,我们发展了正式主义,将这些模型作为多预测空间的次变量进行研究,并引入了一种理论,用于在分阶段树模型这一更普遍的系列中模拟软干预的模型。然后我们考虑了找到其定义方程式的问题,我们得出了一个组合标准,用以确定具有确定性理想的、具有干预性、分阶段的树模型。这一标准,与可分解的DAG模型在相关的分层树木方面的新的定性相结合,在离散的干预式DAG模型中专门使用图形标准。