A context-free language for binary multinomial processing tree models (Q2654150)

Probabilistic binary multinomial processing tree models (BMPTs) are described for categorical data, in which the observed category is a result of occurrence or non-occurrence of some latent binary events. The correspondence between them is described by a binary tree in which the events are the internal nodes while the categories are leafs. An axiomatic theory of BMPTs is developed and a context-free language is proposed for the description of BMPT models. Identifiability of some subclasses of BMPT models and their enumeration are discussed. Some computational algorithms are considered for binary tree plotting with MATLAB software.