Abstract:Probabilistic context-free grammars (PCFGs) and dynamic Bayesian networks (DBNs) are widely used sequence models with complementary strengths and limitations. While PCFGs allow for nested hierarchical dependencies (tree structures), their latent variables (non-terminal symbols) have to be discrete. In contrast, DBNs allow for continuous latent variables, but the dependencies are strictly sequential (chain structure). Therefore, neither can be applied if the latent variables are assumed to be continuous and also to have a nested hierarchical dependency structure. In this paper, we present Recursive Bayesian Networks (RBNs), which generalise and unify PCFGs and DBNs, combining their strengths and containing both as special cases. RBNs define a joint distribution over tree-structured Bayesian networks with discrete or continuous latent variables. The main challenge lies in performing joint inference over the exponential number of possible structures and the continuous variables. We provide two solutions: 1) For arbitrary RBNs, we generalise inside and outside probabilities from PCFGs to the mixed discrete-continuous case, which allows for maximum posterior estimates of the continuous latent variables via gradient descent, while marginalising over network structures. 2) For Gaussian RBNs, we additionally derive an analytic approximation, allowing for robust parameter optimisation and Bayesian inference. The capacity and diverse applications of RBNs are illustrated on two examples: In a quantitative evaluation on synthetic data, we demonstrate and discuss the advantage of RBNs for segmentation and tree induction from noisy sequences, compared to change point detection and hierarchical clustering. In an application to musical data, we approach the unsolved problem of hierarchical music analysis from the raw note level and compare our results to expert annotations.

Pair-copula Bayesian networks

Learning Cluster Causal Diagrams: an Information-Theoretic Approach

The Dual PC Algorithm and the Role of Gaussianity for Structure Learning of Bayesian Networks

Reliability Analysis of Dependent Systems using Copula Bayesian Networks: A Case Study

Bayesian Inference for Regression Copulas

Predicting Binary Outcomes Based on the Pair-Copula Construction

A nonparametric Bayesian approach to copula estimation

Nonparametric estimation of pair-copula constructions with the empirical pair-copula

Bayesian Nonparametric Modelling of Conditional Multidimensional Dependence Structures

A Novel Bayesian Classifier using Copula Functions

Copula Gaussian graphical models and their application to modeling functional disability data

Nonparametric estimation of multivariate copula using empirical bayes method

Copula Approximate Bayesian Computation Using Distribution Random Forests

Parameter estimation for pair-copula constructions

A Bayesian Take on Gaussian Process Networks

Probabilistic Neural Circuits

Bayesian Estimation of Cooccurrence with Dyadic Regression

Recursive Bayesian Networks: Generalising and Unifying Probabilistic Context-Free Grammars and Dynamic Bayesian Networks

Inference, Sampling, and Learning in Copula Cumulative Distribution Networks

Regression Copulas for Multivariate Responses

2-Cats: 2D Copula Approximating Transforms