Latent Variable Model to Infer Food Intake from Multiple Biomarkers

A latent variable model based on factor analytic and mixture of experts models, designed to infer food intake from multiple biomarkers data. The model is framed within a Bayesian hierarchical framework, which provides flexibility to adapt to different biomarker distributions and facilitates inference on food intake from biomarker data alone, along with the associated uncertainty. Details are in D'Angelo, et al. (2020) .

Bayesian Analysis of the Network Autocorrelation Model

The network autocorrelation model (NAM) can be used for studying the degree of social influence regarding an outcome variable based on one or more known networks. The degree of social influence is quantified via the network autocorrelation parameters. In case of a single network, the Bayesian methods of Dittrich, Leenders, and Mulder (2017) and Dittrich, Leenders, and Mulder (2019) are implemented using a normal, flat, or independence Jeffreys prior for the network autocorrelation. In the case of multiple networks, the Bayesian methods of Dittrich, Leenders, and Mulder (2020) are implemented using a multivariate normal prior for the network autocorrelation parameters. Flat priors are implemented for estimating the coefficients. For Bayesian testing of equality and order-constrained hypotheses, the default Bayes factor of Gu, Mulder, and Hoijtink, (2018) is used with the posterior mean and posterior covariance matrix of the NAM parameters based on flat priors as input.

Bayesian Projected Normal Regression Models for Circular Data

Fitting Bayesian multiple and mixed-effect regression models for circular data based on the projected normal distribution. Both continuous and categorical predictors can be included. Sampling from the posterior is performed via an MCMC algorithm. Posterior descriptives of all parameters, model fit statistics and Bayes factors for hypothesis tests for inequality constrained hypotheses are provided. See Cremers, Mulder & Klugkist (2018) and Nuñez-Antonio & Guttiérez-Peña (2014) .

Discriminative Parameter Learning of Bayesian Networks by Differential Evolution

Implements Differential Evolution (DE) to train parameters of Bayesian Networks for optimizing the Conditional Log-Likelihood (Discriminative Learning) instead of the log-likelihood (Generative Learning). Any given Bayesian Network structure encodes assumptions about conditional independencies among the attributes and will result in an error if they do not hold in the data. Such an error includes the classification dimension. The main goal of Discriminative learning is to minimize this type of error. This package provides main variants of differential evolution described in Price & Storn (1996) and recent ones, described in Tanabe & Fukunaga (2014) and Zhang & Sanderson (2009) with adaptation mechanism for factor mutarion and crossover rate.

Bayesian Time Series Modeling with Stan

Fit Bayesian time series models using 'Stan' for full Bayesian inference. A wide range of distributions and models are supported, allowing users to fit Seasonal ARIMA, ARIMAX, Dynamic Harmonic Regression, GARCH, t-student innovation GARCH models, asymmetric GARCH, Random Walks, stochastic volatility models for univariate time series. Prior specifications are flexible and explicitly encourage users to apply prior distributions that actually reflect their beliefs. Model fit can easily be assessed and compared with typical visualization methods, information criteria such as loglik, AIC, BIC WAIC, Bayes factor and leave-one-out cross-validation methods. References: Hyndman (2017) ; Carpenter et al. (2017) .

Inference Tool for Antibody Haplotype

Infers V-D-J haplotypes and gene deletions from AIRR-seq data for Ig and TR chains, based on J, D, or V genes as anchor, by adapting a Bayesian framework. It also calculates a Bayes factor, a number that indicates the certainty level of the inference, for each haplotyped gene. Citation: Gidoni, et al (2019) . Peres and Gidoni, et al (2019) .

Bayesian Information Borrowing with Propensity Score Matching

Hybrid control design is a way to borrow information from external controls to augment concurrent controls in a randomized controlled trial and is expected to overcome the feasibility issue when adequate randomized controlled trials cannot be conducted. A major challenge in the hybrid control design is its inability to eliminate a prior-data conflict caused by systematic imbalances in measured or unmeasured confounding factors between patients in the concurrent treatment/control group and external controls. To prevent the prior-data conflict, a combined use of propensity score matching and Bayesian commensurate prior has been proposed in the context of hybrid control design. The propensity score matching is first performed to guarantee the balance in baseline characteristics, and then the Bayesian commensurate prior is constructed while discounting the information based on the similarity in outcomes between the concurrent and external controls. 'psBayesborrow' is a package to implement the propensity score matching and the Bayesian analysis with commensurate prior, as well as to conduct a simulation study to assess operating characteristics of the hybrid control design, where users can choose design parameters in flexible and straightforward ways depending on their own application.

Hybrid Bayesian Networks Using R and JAGS

Facilities for easy implementation of hybrid Bayesian networks using R. Bayesian networks are directed acyclic graphs representing joint probability distributions, where each node represents a random variable and each edge represents conditionality. The full joint distribution is therefore factorized as a product of conditional densities, where each node is assumed to be independent of its non-descendents given information on its parent nodes. Since exact, closed-form algorithms are computationally burdensome for inference within hybrid networks that contain a combination of continuous and discrete nodes, particle-based approximation techniques like Markov Chain Monte Carlo are popular. We provide a user-friendly interface to constructing these networks and running inference using the 'rjags' package. Econometric analyses (maximum expected utility under competing policies, value of information) involving decision and utility nodes are also supported.

Latent Variable Analysis

Fit a variety of latent variable models, including confirmatory factor analysis, structural equation modeling and latent growth curve models.

Non-Local Alternative Priors in Psychology

Conducts Bayesian Hypothesis tests of a point null hypothesis against a two-sided alternative using Non-local Alternative Prior (NAP) for one- and two-sample z- and t-tests (Pramanik and Johnson, 2022). Under the alternative, the NAP is assumed on the standardized effects size in one-sample tests and on their differences in two-sample tests. The package considers two types of NAP densities: (1) the normal moment prior, and (2) the composite alternative. In fixed design tests, the functions calculate the Bayes factors and the expected weight of evidence for varied effect size and sample size. The package also provides a sequential testing framework using the Sequential Bayes Factor (SBF) design. The functions calculate the operating characteristics (OC) and the average sample number (ASN), and also conducts sequential tests for a sequentially observed data.