Infinite mixtures for multi-relational categorical data
نویسندگان
چکیده
Large relational datasets are prevalent in many fields. We propose an unsupervised component model for relational data, i.e., for heterogeneous collections of categorical co-occurrences. The co-occurrences can be dyadic or n-adic, and over the same or different categorical variables. Graphs are a special case, as collections of dyadic cooccurrences (edges) over a set of vertices. The model is simple, with only one latent variable. This allows wide applicability as long as a global latent component solution is preferred, and the generative process fits the application. Estimation with a collapsed Gibbs sampler is straightforward. We demostrate the model with graphs enriched with multinomial vertex properties, or more conceretely, with two sets of scientific papers, with both content and citation information available.
منابع مشابه
A new solving approach for fuzzy multi-objective programming problem in uncertainty conditions by using semi-infinite linear programing
In practice, there are many problems which decision parameters are fuzzy numbers, and some kind of this problems are formulated as either possibilitic programming or multi-objective programming methods. In this paper, we consider a multi-objective programming problem with fuzzy data in constraints and introduce a new approach for solving these problems base on a combination of the multi-objecti...
متن کاملMeasurement of Selected Thermophysical and Viscosity of Binary Mixtures of Nitromethane with 2-Propanol, 2-Butanol and 2-Pentanol
The density and viscosity values at T= (293.15 to 323.15) K for binary mixtures of nitromethane with 2-propanol, 2-butanol and 2-pentanol were measured over the entire composition range. From the experimental data, excess molar volumes, thermal expansion coefficients, excess thermal expansions, partial molar volumes at infinite dilution and deviations in viscosity were calculated. The results w...
متن کاملSuitability of Programming Languages for Categorical Databases
Earlier work has defined a formalism for object-relational databases, based on category theory, using the concepts of multi-level mappings and products. A prototype implementation of this database model, known as the Product Model, is to be developed, but there are known problems in languages providing rich enough structures in which to produce a categorical model. This paper examines five well...
متن کاملThe Infinite Mixture of Infinite Gaussian Mixtures for Clustering Data Sets with Multi-mode and Rare Clusters Supplementary Material
متن کامل
Covering Morphisms in Categories of Relational Algebras
In this paper we use Janelidze’s approach to the classical theory of topological coverings via categorical Galois theory to study coverings in categories of relational algebras. Moreover, we present characterizations of effective descent morphisms in the categories of M ordered sets and of multi-ordered sets.
متن کامل