Rafael Cabral

Rafael Cabral

Statistician / Data Scientist

Publications

Fitting latent non-Gaussian models using variational Bayes and Laplace approximations

Published in arXiv, 2022

In this paper, we derive variational Bayes algorithms for fast and scalable inference of latent non-Gaussian models. To facilitate Bayesian inference, we introduce the ngvb package, where LGMs implemented in R-INLA can be easily extended to LnGMs by adding a single line of code.

Recommended citation: Cabral, R., Bolin, D. and Rue, H. (2022). "Fitting latent non-Gaussian models using variational Bayes and Laplace approximations. " arXiv preprint.

Controlling the flexibility of non‑Gaussian processes through shrinkage priors

Published in Bayesian Analysis, 2022

Inferential procedures tend to overestimate the degree of non-Gaussianity in the data and therefore we construct priors that contract the model towards Gaussianity. In our venture to derive sensible priors, we also propose a new intuitive parameterization of the non-Gaussian models and discuss how to implement them efficiently in Stan.

Recommended citation: Cabral, R., Bolin, D. and Rue, H. (2022). "Controlling the flexibility of non‑Gaussian processes through shrinkage priors." Bayesian Analysis.

A price model with finitely many agents

Published in Bulletin of the Portuguese Mathematical Society, 2019

We propose a price-formation model, with a population consisting of a finite number of agents storing and trading a commodity. We formulate our problem as an N-player dynamic game with a market-clearing condition. Subsequently, we show how to recast our game as an optimization problem for the overall trading cost. We show the existence of a solution using the direct method in the calculus of variations.

Recommended citation: Alharbi, A., Bakaryan, T., Cabral, R., Campi, S., Christoffersen, N., Colusso, P., Costa, O., Duisembay, S., Ferreira, R., Gomes, D.A. and Guo, S., (2019). "A price model with finitely many agents." Bulletin of the Portuguese Mathematical Society.

Space‑time trends and dependence of precipitation extremes in north‑western Germany

Published in Environmetrics, 2019

We propose a new approach for evaluating temporal trends and spatial homogeneity in extremes accounting also for spatial dependence. Wwo novel test statistics provide a way to assess space–time inhomogeneities. Finally, we propose a procedure to achieve stationarity in space and time and evaluate residual extremal dependence over space through a variogram analysis that includes anisotropy.

Recommended citation: Cabral, R, Ferreira, A, Friederichs, (2019). "Space‑time trends and dependence of precipitation extremes in north‑western Germany." Environmetrics.