arXiv subjects

Brian J. Reich

Publications and source records attributed to Brian J. Reich.

At least 55 records · Page 3Linked to original sources

arXiv2020

Bayesian Regression Using a Prior on the Model Fit: The R2-D2 Shrinkage Prior

Prior distributions for high-dimensional linear regression require specifying a joint distribution for the unobserved regression coefficients, which is inherently difficult. We instead propose a new class of shrinkage priors for linear regression via specifying a prior first on the model fit, in particular, the coefficient of determination, and then distributing through to the coefficients in a novel way. The proposed method compares favourably to previous approaches in terms of both concentration around the origin and tail behavior, which leads to improved performance both in posterior contraction and in empirical performance. The limiting behavior of the proposed prior is $1/x$, both around the origin and in the tails. This behavior is optimal in the sense that it simultaneously lies on the boundary of being an improper prior both in the tails and around the origin. None of the existing shrinkage priors obtain this behavior in both regions simultaneously. We also demonstrate that our proposed prior leads to the same near-minimax posterior contraction rate as the spike-and-slab prior.

stat.ME↗

arXiv2020

Spatial shrinkage via the product independent Gaussian process prior

We study the problem of sparse signal detection on a spatial domain. We propose a novel approach to model continuous signals that are sparse and piecewise smooth as product of independent Gaussian processes (PING) with a smooth covariance kernel. The smoothness of the PING process is ensured by the smoothness of the covariance kernels of Gaussian components in the product, and sparsity is controlled by the number of components. The bivariate kurtosis of the PING process shows more components in the product results in thicker tail and sharper peak at zero. The simulation results demonstrate the improvement in estimation using the PING prior over Gaussian process (GP) prior for different image regressions. We apply our method to a longitudinal MRI dataset to detect the regions that are affected by multiple sclerosis (MS) in the greatest magnitude through an image-on-scalar regression model. Due to huge dimensionality of these images, we transform the data into the spectral domain and develop methods to conduct computation in this domain. In our MS imaging study, the estimates from the PING model are more informative than those from the GP model.

stat.ME↗

arXiv2020

Accounting for Location Measurement Error in Imaging Data with Application to Atomic Resolution Images of Crystalline Materials

Scientists use imaging to identify objects of interest and infer properties of these objects. The locations of these objects are often measured with error, which when ignored leads to biased parameter estimates and inflated variance. Current measurement error methods require an estimate or knowledge of the measurement error variance to correct these estimates, which may not be available. Instead, we create a spatial Bayesian hierarchical model that treats the locations as parameters, it using the image itself to incorporate positional uncertainty. We lower the computational burden by approximating the likelihood using a non-contiguous block design around the object locations. We apply this model in a materials science setting to study the relationship between the chemistry and displacement of hundreds of atom columns in crystal structures directly imaged via scanning transmission electron microscopy. Greater knowledge of this relationship can lead to engineering materials with improved properties of interest. We find strong evidence of a negative relationship between atom column displacement and the intensity of neighboring atom columns, which is related to the local chemistry. A simulation study shows our method corrects the bias in the parameter of interest and drastically improves coverage in high noise scenarios compared to non-measurement error models.

stat.AP↗

arXiv2020

A Bayesian semi-parametric hybrid model for spatial extremes with unknown dependence structure

The max-stable process is an asymptotically justified model for spatial extremes. In particular, we focus on the hierarchical extreme-value process (HEVP), which is a particular max-stable process that is conducive to Bayesian computing. The HEVP and all max-stable process models are parametric and impose strong assumptions including that all marginal distributions belong to the generalized extreme value family and that nearby sites are asymptotically dependent. We generalize the HEVP by relaxing these assumptions to provide a wider class of marginal distributions via a Dirichlet process prior for the spatial random effects distribution. In addition, we present a hybrid max-mixture model that combines the strengths of the parametric and semi-parametric models. We show that this versatile max-mixture model accommodates both asymptotic independence and dependence and can be fit using standard Markov chain Monte Carlo algorithms. The utility of our model is evaluated in Monte Carlo simulation studies and application to Netherlands wind gust data.

stat.ME↗

arXiv2020

A spatial causal analysis of wildland fire-contributed PM2.5 using numerical model output

Wildland fire smoke contains hazardous levels of fine particulate matter PM2.5, a pollutant shown to adversely effect health. Estimating fire attributable PM2.5 concentrations is key to quantifying the impact on air quality and subsequent health burden. This is a challenging problem since only total PM2.5 is measured at monitoring stations and both fire-attributable PM2.5 and PM2.5 from all other sources are correlated in space and time. We propose a framework for estimating fire-contributed PM2.5 and PM2.5 from all other sources using a novel causal inference framework and bias-adjusted chemical model representations of PM2.5 under counterfactual scenarios. The chemical model representation of PM2.5 for this analysis is simulated using Community Multi-Scale Air Quality Modeling System (CMAQ), run with and without fire emissions across the contiguous U.S. for the 2008-2012 wildfire seasons. The CMAQ output is calibrated with observations from monitoring sites for the same spatial domain and time period. We use a Bayesian model that accounts for spatial variation to estimate the effect of wildland fires on PM2.5 and state assumptions under which the estimate has a valid causal interpretation. Our results include estimates of absolute, relative and cumulative contributions of wildfire smoke to PM2.5 for the contiguous U.S. Additionally, we compute the health burden associated with the PM2.5 attributable to wildfire smoke.

stat.ME↗

arXiv2020

A spatiotemporal recommendation engine for malaria control

Malaria is an infectious disease affecting a large population across the world, and interventions need to be efficiently applied to reduce the burden of malaria. We develop a framework to help policy-makers decide how to allocate limited resources in realtime for malaria control. We formalize a policy for the resource allocation as a sequence of decisions, one per intervention decision, that map up-to-date disease related information to a resource allocation. An optimal policy must control the spread of the disease while being interpretable and viewed as equitable to stakeholders. We construct an interpretable class of resource allocation policies that can accommodate allocation of resources residing in a continuous domain, and combine a hierarchical Bayesian spatiotemporal model for disease transmission with a policy-search algorithm to estimate an optimal policy for resource allocation within the pre-specified class. The estimated optimal policy under the proposed framework improves the cumulative long-term outcome compared with naive approaches in both simulation experiments and application to malaria interventions in the Democratic Republic of the Congo.

stat.AP↗

arXiv2020

Constrained Bayesian Nonparametric Regression for Grain Boundary Energy Predictions

Grain boundary (GB) energy is a fundamental property that affects the form of grain boundary and plays an important role to unveil the behavior of polycrystalline materials. With a better understanding of grain boundary energy distribution (GBED), we can produce more durable and efficient materials that will further improve productivity and reduce loss. The lack of robust GB structure-property relationships still remains one of the biggest obstacles towards developing true bottom-up models for the behavior of polycrystalline materials. Progress has been slow because of the inherent complexity associated with the structure of interfaces and the vast five-dimensional configurational space in which they reside. Estimating the GBED is challenging from a statistical perspective because there are not direct measurements on the grain boundary energy. We only have indirect information in the form of an unidentifiable homogeneous set of linear equations. In this paper, we propose a new statistical model to determine the GBED from the microstructures of polycrystalline materials. We apply spline-based regression with constraints to successfully recover the GB energy surface. Hamiltonian Monte Carlo and Gibbs sampling are used for computation and model fitting. Compared with conventional methods, our method not only gives more accurate predictions but also provides prediction uncertainties.

physics.comp-ph↗

arXiv2019

Statistical downscaling with spatial misalignment: Application to wildland fire PM$_{2.5}$ concentration forecasting

Fine particulate matter, PM$_{2.5}$, has been documented to have adverse health effects and wildland fires are a major contributor to PM$_{2.5}$ air pollution in the US. Forecasters use numerical models to predict PM$_{2.5}$ concentrations to warn the public of impending health risk. Statistical methods are needed to calibrate the numerical model forecast using monitor data to reduce bias and quantify uncertainty. Typical model calibration techniques do not allow for errors due to misalignment of geographic locations. We propose a spatiotemporal downscaling methodology that uses image registration techniques to identify the spatial misalignment and accounts for and corrects the bias produced by such warping. Our model is fitted in a Bayesian framework to provide uncertainty quantification of the misalignment and other sources of error. We apply this method to different simulated data sets and show enhanced performance of the method in the presence of spatial misalignment. Finally, we apply the method to a large fire in Washington state and show that the proposed method provides more realistic uncertainty quantification than standard methods.

stat.AP↗

arXiv2019

Geostatistical Modeling of Positive Definite Matrices: An Application to Diffusion Tensor Imaging

Geostatistical modeling for continuous point-referenced data has been extensively applied to neuroimaging because it produces efficient and valid statistical inference. However, diffusion tensor imaging (DTI), a neuroimaging characterizing the brain structure produces a positive definite (p.d.) matrix for each voxel. Current geostatistical modeling has not been extended to p.d. matrices because introducing spatial dependence among positive definite matrices properly is challenging. In this paper, we use the spatial Wishart process, a spatial stochastic process (random field) where each p.d. matrix-variate marginally follows a Wishart distribution, and spatial dependence between random matrices is induced by latent Gaussian processes. This process is valid on an uncountable collection of spatial locations and is almost surely continuous, leading to a reasonable means of modeling spatial dependence. Motivated by a DTI dataset of cocaine users, we propose a spatial matrix-variate regression model based on the spatial Wishart process. A problematic issue is that the spatial Wishart process has no closed-form density function. Hence, we propose approximation methods to obtain a feasible working model. A local likelihood approximation method is also applied to achieve fast computation. The simulation studies and real data analysis demonstrate that the working model produces reliable inference and improved performance compared to other methods.

stat.ME↗

arXiv2019

Mean-dependent nonstationary spatial models

Nonstationarity is a major challenge in analyzing spatial data. For example, daily precipitation measurements may have increased variability and decreased spatial smoothness in areas with high mean rainfall. Common nonstationary covariance models introduce parameters specific to each location, giving a highly-parameterized model which is difficult to fit. We develop a nonstationary spatial model that uses the mean to determine the covariance in a region, resulting in a far simpler, albeit more specialized, model. We explore inferential and predictive properties of the model under various simulated data situations. We show that this model in certain circumstances improves predictions compared to a standard stationary spatial model. We further propose a computationally efficient approximation that has comparable predictive accuracy. We also develop a test for nonstationary data and show it reliably identifies nonstationarity. We apply these methods to daily precipitation in Puerto Rico.

stat.ME↗

arXiv2019

Global forensic geolocation with deep neural networks

An important problem in forensic analyses is identifying the provenance of materials at a crime scene, such as biological material on a piece of clothing. This procedure, known as geolocation, is conventionally guided by expert knowledge of the biological evidence and therefore tends to be application-specific, labor-intensive, and subjective. Purely data-driven methods have yet to be fully realized due in part to the lack of a sufficiently rich data source. However, high-throughput sequencing technologies are able to identify tens of thousands of microbial taxa using DNA recovered from a single swab collected from nearly any object or surface. We present a new algorithm for geolocation that aggregates over an ensemble of deep neural network classifiers trained on randomly-generated Voronoi partitions of a spatial domain. We apply the algorithm to fungi present in each of 1300 dust samples collected across the continental United States and then to a global dataset of dust samples from 28 countries. Our algorithm makes remarkably good point predictions with more than half of the geolocation errors under 100 kilometers for the continental analysis and nearly 90% classification accuracy of a sample's country of origin for the global analysis. We suggest that the effectiveness of this model sets the stage for a new, quantitative approach to forensic geolocation.

stat.ML↗

arXiv2019

A Spatial Bayesian Semiparametric Mixture Model for Positive Definite Matrices with Applications to Diffusion Tensor Imaging

Diffusion tensor imaging (DTI) is a popular magnetic resonance imaging technique used to characterize microstructural changes in the brain. DTI studies quantify the diffusion of water molecules in a voxel using an estimated 3x3 symmetric positive definite diffusion tensor matrix. Statistical analysis of DTI data is challenging because the data are positive definite matrices. Matrix-variate information is often summarized by a univariate quantity, such as the fractional anisotropy (FA), leading to a loss of information. Furthermore, DTI analyses often ignore the spatial association of neighboring voxels, which can lead to imprecise estimates. Although the spatial modeling literature is abundant, modeling spatially dependent positive definite matrices is challenging. To mitigate these issues, we propose a matrix-variate Bayesian semiparametric mixture model, where the positive definite matrices are distributed as a mixture of inverse Wishart distributions with the spatial dependence captured by a Markov model for the mixture component labels. Conjugacy and the double Metropolis-Hastings algorithm result in fast and elegant Bayesian computing. Our simulation study shows that the proposed method is more powerful than non-spatial methods. We also apply the proposed method to investigate the effect of cocaine use on brain structure. The contribution of our work is to provide a novel statistical inference tool for DTI analysis by extending spatial statistics to matrix-variate data.

stat.ME↗

arXiv2019

Deep Distribution Regression

Due to their flexibility and predictive performance, machine-learning based regression methods have become an important tool for predictive modeling and forecasting. However, most methods focus on estimating the conditional mean or specific quantiles of the target quantity and do not provide the full conditional distribution, which contains uncertainty information that might be crucial for decision making. In this article, we provide a general solution by transforming a conditional distribution estimation problem into a constrained multi-class classification problem, in which tools such as deep neural networks. We propose a novel joint binary cross-entropy loss function to accomplish this goal. We demonstrate its performance in various simulation studies comparing to state-of-the-art competing methods. Additionally, our method shows improved accuracy in a probabilistic solar energy forecasting problem.

stat.ML↗

arXiv2019

Sequential Optimization in Locally Important Dimensions

Optimizing an expensive, black-box function $f(\cdot)$ is challenging when its input space is high-dimensional. Sequential design frameworks first model $f(\cdot)$ with a surrogate function and then optimize an acquisition function to determine input settings to evaluate next. Optimization of both $f(\cdot)$ and the acquisition function benefit from effective dimension reduction. Global variable selection detects and removes input variables that do not affect $f(\cdot)$ across the input space. Further dimension reduction may be possible if we consider local variable selection around the current optimum estimate. We develop a sequential design algorithm called Sequential Optimization in Locally Important Dimensions (SOLID) that incorporates global and local variable selection to optimize a continuous, differentiable function. SOLID performs local variable selection by comparing the surrogate's predictions in a localized region around the estimated optimum with the $p$ alternative predictions made by removing each input variable. The search space of the acquisition function is further restricted to focus only on the variables that are deemed locally active, leading to greater emphasis on refining the surrogate model in locally active dimensions. A simulation study across three test functions and an application to the Sarcos robot dataset show that SOLID outperforms conventional approaches.

stat.ME↗

arXiv2018

A multivariate spatial skew-t process for joint modeling of extreme precipitation indexes

To study trends in extreme precipitation across US over the years 1951-2017, we consider 10 climate indexes that represent extreme precipitation, such as annual maximum of daily precipitation, annual maximum of consecutive 5-day average precipitation, which exhibit spatial correlation as well as mutual dependence. We consider the gridded data, produced by the CLIMDEX project (http://www.climdex.org/gewocs.html), constructed using daily precipitation data. In this paper, we propose a multivariate spatial skew-t process for joint modeling of extreme precipitation indexes and discuss its theoretical properties. The model framework allows Bayesian inference while maintaining a computational time that is competitive with common multivariate geostatistical approaches. In a numerical study, we find that the proposed model outperforms multivariate spatial Gaussian processes, multivariate spatial t-processes including their univariate alternatives in terms of various model selection criteria. We apply the proposed model to estimate the average decadal change in the extreme precipitation indexes throughout the United States and find several significant local changes.

stat.AP↗

arXiv2018

Bayesian Nonparametric Policy Search with Application to Periodontal Recall Intervals

Tooth loss from periodontal disease is a major public health burden in the United States. Standard clinical practice is to recommend a dental visit every six months; however, this practice is not evidence-based, and poor dental outcomes and increasing dental insurance premiums indicate room for improvement. We consider a tailored approach that recommends recall time based on patient characteristics and medical history to minimize disease progression without increasing resource expenditures. We formalize this method as a dynamic treatment regime which comprises a sequence of decisions, one per stage of intervention, that follow a decision rule which maps current patient information to a recommendation for their next visit time. The dynamics of periodontal health, visit frequency, and patient compliance are complex, yet the estimated optimal regime must be interpretable to domain experts if it is to be integrated into clinical practice. We combine non-parametric Bayesian dynamics modeling with policy-search algorithms to estimate the optimal dynamic treatment regime within an interpretable class of regimes. Both simulation experiments and application to a rich database of electronic dental records from the HealthPartners HMO shows that our proposed method leads to better dental health without increasing the average recommended recall time relative to competing methods.

stat.AP↗

arXiv2018

Exploration and inference in spatial extremes using empirical basis functions

Statistical methods for inference on spatial extremes of large datasets are yet to be developed. Motivated by standard dimension reduction techniques used in spatial statistics, we propose an approach based on empirical basis functions to explore and model spatial extremal dependence. Based on a low-rank max-stable model we propose a data-driven approach to estimate meaningful basis functions using empirical pairwise extremal coefficients. These spatial empirical basis functions can be used to visualize the main trends in extremal dependence. In addition to exploratory analysis, we describe how these functions can be used in a Bayesian hierarchical model to model spatial extremes of large datasets. We illustrate our methods on extreme precipitations in eastern U.S.

stat.ME↗

arXiv2018

A nonparametric spatial test to identify factors that shape a microbiome

The advent of high-throughput sequencing technologies has made data from DNA material readily available, leading to a surge of microbiome-related research establishing links between markers of microbiome health and specific outcomes. However, to harness the power of microbial communities we must understand not only how they affect us, but also how they can be influenced to improve outcomes. This area has been dominated by methods that reduce community composition to summary metrics, which can fail to fully exploit the complexity of community data. Recently, methods have been developed to model the abundance of taxa in a community, but they can be computationally intensive and do not account for spatial effects underlying microbial settlement. These spatial effects are particularly relevant in the microbiome setting because we expect communities that are close together to be more similar than those that are far apart. In this paper, we propose a flexible Bayesian spike-and-slab variable selection model for presence-absence indicators that accounts for spatial dependence and cross-dependence between taxa while reducing dimensionality in both directions. We show by simulation that in the presence of spatial dependence, popular distance-based hypothesis testing methods fail to preserve their advertised size, and the proposed method improves variable selection. Finally, we present an application of our method to an indoor fungal community found with homes across the contiguous United States.

stat.ME↗