José Carlos Ferreira da Rocha

Probabilistic logic with independence

This paper investigates probabilistic logics endowed with independence relations. We review propositional probabilistic languages without and with independence. We then consider graph-theoretic representations for propositional probabilistic logic with independence; complexity is analyzed, algorithms are derived, and examples are discussed. Finally, we examine a restricted first-order probabilistic logic that generalizes relational Bayesian networks.

Inference in credal networks: branch-and-bound methods and the A/R+ algorithm

A credal network is a graphical representation for a set of joint probability distributions. In this paper we discuss algorithms for exact and approximate inferences in credal networks. We propose a branch-and-bound framework for inference, and focus on inferences for polytree-shaped networks. We also propose a new algorithm, A/R+, for outer approximations in polytree-shaped credal networks.

Evidence Propagation in Credal Networks: An Exact Algorithm Based on Separately Specified Sets of Probability

Probabilistic models and graph-based independence languages have often been combined in artificial intelligence research. The Bayesian network formalism is probably the best example of this type of association. In this article we focus on graphical structures that associate graphs with sets of probability measures -- the result is referred to as a credal network. We describe credal networks and review an algorithm for evidential reasoning that we have recently developed. The algorithm substantially simplifies the computation of upper and lower probabilities by exploiting an independence assumption (strong independence) and a representation based on separately specified sets of probability measures. The algorithm is particularly efficient when applied to polytree structures. We then discuss a strategy for approximate reasoning in multi-connected networks, based on conditioning.

Ribopeaks: a web tool for bacterial classification through m/z data from ribosomal proteins

Summary MALDI-TOF MS is a rapid, sensitive and economic tool for bacterial identification. Highly abundant bacterial proteins are detected by this technique, including ribosomal proteins (r-protein), and the generated mass spectra are compared with a MALDI-TOF MS spectra database. Currently, it allows mainly the classification of clinical bacteria due to the limited number of environmental bacteria included in the spectra database. We present a wide-ranging bacterium classifier tool, called Ribopeaks, which was created based on r-protein data from the Genbank. The Ribopeaks database has more than 28 500 bacterial taxonomic records. It compares the incoming m/z data from MALDI-TOF MS analysis with models stored in the Ribopeaks database created by machine learning and then taxonomically classifies the bacteria. Availability and implementation The software is available at http://www.ribopeaks.com. Supplementary information Supplementary data are available at Bioinformatics online.

Probabilistic logic with strong independence

This papers investigates the manipulation of statements of strong independence in probabilistic logic. Inference methods based on polynomial programming are presented for strong independence, both for unconditional and conditional cases. We also consider graph-theoretic representations, where each node in a graph is associated with a Boolean variable and edges carry a Markov condition. The resulting model generalizes Bayesian networks, allowing probabilistic assessments and logical constraints to be mixed.