Accurate prediction of an insurer’s outstanding liabilities is crucial for maintaining the financial health of the insurance sector. We aim to develop a statistical model for insurers to dynamically forecast unpaid losses by leveraging the granular transaction data on individual claims. The liability cash flow from a single insurance claim is determined by an event process that describes the recurrences of payments, a payment process that generates a sequence of payment amounts, and a settlement process that terminates both the event and payment processes. More importantly, the three components are dependent on one another, which enables the dynamic prediction of an insurer’s outstanding liability. We introduce a copula-based point process framework to model the recurrent events of payment transactions from an insurance claim, where the longitudinal payment amounts and the time-to-settlement outcome are formulated as the marks and the terminal event of the counting process, respectively. The dependencies among the three components are characterized using the method of pair copula constructions. We further develop a stagewise strategy for parameter estimation and illustrate its desirable properties with numerical experiments.

In the application we consider a portfolio of property insurance claims for building and contents coverage obtained from a commercial property insurance provider, where we find intriguing dependence patterns among the three components. The superior dynamic prediction performance of the proposed joint model enhances the insurer’s decision-making in claims reserving and risk financing operations.

Optimal sampling strategies are critical for surveys of deeper coral reef and shoal systems due to the significant cost of accessing and field sampling these remote and poorly understood ecosystems. Additionally, well-established standard diver-based sampling techniques used in shallow reef systems are not feasible at greater depths. In this study, we develop a Bayesian design strategy to optimise sampling for a shoal deep reef system using three years of pilot data. Bayesian designs are typically found by maximising the expectation of a utility function with respect to the joint distribution of the parameters and the response conditional on an assumed statistical model. Unfortunately, specifying such a model a priori is difficult, as knowledge of the data-generating process is typically incomplete. To overcome this, our approach focuses on finding Bayesian designs that are robust to unknown model uncertainty. We achieve this by couching the specified model within a generalised additive modelling framework and formulating prior information that allows the additive component to capture discrepancies between what is assumed and the underlying data-generating process. The motivation for this is to enable Bayesian designs to be found under epistemic model uncertainty; a highly desirable property of Bayesian designs. Initially, we demonstrate our approach with an exemplar design problem, deriving a theoretical result to explore the properties of optimal designs. We then apply this approach to design future monitoring of submerged shoals off the north-west coast of Australia to improve current monitoring practices.

Accelerometry data enables scientists to extract personal digital features useful in precision health decision making. Existing analytic methods often begin with discretizing physical activity (PA) counts into activity categories via fixed cutoffs; however, the cutoffs are validated under restricted settings and cannot be generalized across studies. Here we develop a data-driven approach to overcome this bottleneck in the analysis of PA data in which we holistically summarize an individual’s PA profile using occupation-time curves that describe the percentage of time spent at or above a continuum of activity levels. The resulting functional curve is informative to capture time-course individual variability of PA. We investigate functional analytics under an ${\mathit{L}}_0$ regularization approach, which handles highly correlated micro-activity windows that serve as predictors in a scalar-on-function regression model. We develop a new one-step method that simultaneously conducts fusion via change-point detection and parameter estimation through a new ${\mathit{L}}_0$ constraint formulation, which is evaluated via simulation experiments and a data analysis assessing the influence of PA on biological aging.

Causal inference with spatial environmental data is often challenging due to the presence of interference: outcomes for observational units depend on some combination of local and nonlocal treatment. This is especially relevant when estimating the effect of power plant emissions controls on population health, as pollution exposure is dictated by: (i) the location of point-source emissions as well as (ii) the transport of pollutants across space via dynamic physical-chemical processes. In this work we estimate the effectiveness of air quality interventions at coal-fired power plants in reducing two adverse health outcomes in Texas in 2016: pediatric asthma ED visits and Medicare all-cause mortality. We develop methods for causal inference with interference when the underlying network structure is not known with certainty and instead must be estimated from ancillary data. Notably, uncertainty in the interference structure is propagated to the resulting causal effect estimates. We offer a Bayesian, spatial mechanistic model for the interference mapping, which we combine with a flexible nonparametric outcome model to marginalize estimates of causal effects over uncertainty in the structure of interference. Our analysis finds some evidence that emissions controls at upwind power plants reduce asthma ED visits and all-cause mortality; however, accounting for uncertainty in the interference renders the results largely inconclusive.

Aging changes brain functions and structures in a downward trajectory and consequently leads to a decline in neurocognitive performance. Our research is motivated by understanding whether and to what extent the age-effect on cognitive decline can be explained by neuroimaging measures. We consider a new mediation model with age as an independent variable, while treating neuroimaging data and cognitive function as the multiple mediators and outcome, respectively. Given that the brain is the primary organ responsible for cognitive function, it is neurobiologically intuitive that the age-related decline in cognition is largely mediated through neuroimaging measures. Additionally, cognitive function is localized to certain regions of the brain rather than being a function of the entire brain. Taking these factors into account, we propose a novel mediation model with multiple mediators that aims to maximally uncover the mediation pathway while simultaneously identifying active neuroimaging mediators by imposing an ${\ell }_1$ penalty and ${\ell }_2$ constraint. We develop a computationally efficient algorithm to handle the nonconvex optimization problem of penalized mediation proportion maximization. We apply our method to a data example of 37,441 participants of UK Biobank with cortical gray-matter thickness and white-matter integrity measures and cognitive performance scores. Our results show that the mediation effect of brain-imaging variables can explain 97% of age-related cognitive decline.

To optimize personalized treatment strategies and extend patients’ survival times, it is critical to accurately predict patients’ prognoses at all stages, from disease diagnosis to follow-up visits. The longitudinal biomarker measurements during visits are essential for this prediction purpose. Patients’ ultimate concerns are cure and survival. However, in many situations there is no clear biomarker indicator for cure. We propose a comprehensive joint model of longitudinal and survival data and a landmark cure model, incorporating proportions of potentially cured patients. The survival distributions in the joint and landmark models are specified through flexible hazard functions with the proportional hazards as a special case, allowing other patterns such as crossing hazard and survival functions. Formulas are provided for predicting each individual’s probabilities of future cure and survival at any time point based on his or her current biomarker history. Simulations show that, with these comprehensive and flexible properties, the proposed cure models outperform standard cure models in terms of predictive performance, measured by the time-dependent area under the curve of receiver operating characteristic, Brier score, and integrated Brier score. The use and advantages of the proposed models are illustrated by their application to a study of patients with chronic myeloid leukemia.

Risk assessment for extreme events requires accurate estimation of high quantiles that go beyond the range of historical observations. When the risk depends on the values of observed predictors, regression techniques are used to interpolate in the predictor space. We propose the EQRN model that combines tools from neural networks and extreme value theory into a method capable of extrapolation in the presence of complex predictor dependence. Neural networks can naturally incorporate additional structure in the data. We develop a recurrent version of EQRN that is able to capture complex sequential dependence in time series. We apply this method to forecast flood risk in the Swiss Aare catchment. It exploits information from multiple covariates in space and time to provide one-day-ahead predictions of return levels and exceedance probabilities. This output complements the static return level from a traditional extreme value analysis, and the predictions are able to adapt to distributional shifts as experienced in a changing climate. Our model can help authorities to manage flooding more effectively and to minimize their disastrous impacts through early warning systems.

Predictive uncertainty quantification is crucial for reliable decision-making in various applied domains. Bayesian neural networks offer a powerful framework for this task. However, defining meaningful priors and ensuring computational efficiency remain significant challenges, especially for complex real-world applications. This paper addresses these challenges by proposing a novel neural adaptive empirical Bayes (NA-EB) framework. NA-EB leverages a class of implicit generative priors derived from low-dimensional distributions. This allows for efficient handling of complex data structures and effective capture of underlying relationships in real-world datasets. The proposed NA-EB framework combines variational inference with a gradient ascent algorithm. This enables simultaneous hyperparameter selection and approximation of the posterior distribution, leading to improved computational efficiency. We establish the theoretical foundation of the framework through posterior and classification consistency. We demonstrate the practical applications of our framework through extensive evaluations on a variety of tasks, including the two-spiral problem, regression, 10 UCI datasets, and image classification tasks on both MNIST and CIFAR-10 datasets. The results of our experiments highlight the superiority of our proposed framework over existing methods, such as sparse variational Bayesian and generative models, in terms of prediction accuracy and uncertainty quantification.

Personal credits have always been a hot topic in the society. Among all of them, the evaluation of default risk is particularly concerned since robust estimation, based on personal information, can both help needy individuals to get loans and financial institutions to avoid losses. So far, there have been no good solutions due to limited data, especially default information. With the advent of the era of big data, it is possible to improve the effectiveness of estimates by using auxiliary information from external studies or public domains. However, the individual-level data can not be gained directly because of the emphasis on data privacy; that is, only some summarized statistics with auxiliary information are allowed to be shared. To effectively utilize external integrated auxiliary information to improve the accuracy of default risk estimation, this paper introduces a unified auxiliary information framework, which is referred as enhanced GEE method, to effectively incorporate various external summary results by employing the generalized estimating equations (GEE) approach and augmenting a weighted logarithm of confidence density on GEE function. We establish asymptotic properties for the new method and prove that it can achieve the gain of statistical efficiency compared to the study-specific estimator without any auxiliary information. Besides, a low-cost Map-Reduce procedure for the distributed statistical inference of enhanced GEE method in big data is developed that can achieve the same efficiency as the oracle enhanced GEE approach under mild condition. This method is demonstrated by an application to predict the loan default risk of bank customers in Shanghai and shown to be more effective and reliable compared with the method based on the own data only. Furthermore, the superiorities of our approach, especially the construction of the tighter confidence intervals, are also illustrated with extensive simulation studies and a real personal default risk case.

Since 2014, states in the U.S. can choose whether to adopt Medicaid expansion as part of the Affordable Care Act (ACA), relaxing eligibility requirements. This heterogeneity in policy adoption between states raises the question—would there be a difference in health outcomes for states that have not expanded insurance access if they did expand Medicaid eligibility? In this study we estimate the effect of ACA Medicaid expansion on county-level all-cause mortality in the U.S. in 2014 overall and for subgroups relevant to the racial politics surrounding the ACA. We bring a causal approach to this challenge which emphasizes observational study design, including prespecifying all analyses, matching counties on pretreatment covariates, and employing design-based inference.

A challenge facing analyses like this one is gaining access to mortality outcomes, as statistical agencies in the U.S. and elsewhere suppress cell counts of 10 or fewer in public use data. We develop a rank-sum test statistic accommodating outcomes that are coarsened in this way and that lends itself to design-based inference with county-aggregated data. As applied to impact analysis of the ACA’s Medicaid expansion, the proposed method’s inferences from coarsened, publicly available data are substantively the same as those that would be drawn from the complete, restricted-access data.

Count data at surveyed sites are an important monitoring tool for several species around the world. However, the raw count data are an underestimate of the size of the monitored population at any one time, as individuals can temporarily leave the site (temporary emigration, TE) and because the probability of detection of individuals, even when using the site, is typically much lower than one (observation error). In this paper we develop a novel modelling framework for estimating population size, from count data, while accounting for both TE and observation error. Our framework builds on the popular class of N-mixture models but extends them in a number of ways. Specifically, we introduce two model classes for TE, a parametric, which relies on temporal models, and a nonparametric, which relies on Dirichlet process mixture models. Both model classes give rise to interesting ecological interpretations of the TE pattern while being parsimonious in terms of the number of parameters required to model the pattern. When accounting for observation error, we use mixed-effects models and implement an efficient Bayesian variable selection algorithm for identifying important predictors for the probability of detection. We demonstrate our new modelling framework using an extensive simulation study, which highlights the importance of using mixed-effects models for the probability of detection and illustrates the performance of the model when estimating population size and underlying TE patterns. We also assess the ability of the corresponding variable selection algorithm to identify important predictors under different scenarios for observation error and its corresponding model. When fitted to two motivating data sets of parrots counted at their roosts, our results provide new insights into how each species uses the roost throughout the year, on changes in population size between and within years, and on observation error.

Even with daily case counts, the true scope of the COVID-19 pandemic in Canada is unknown due to undetected cases. We develop a novel multisite disease analytics model which estimates undetected cases using discrete-valued multivariate time series in the framework of Bayesian hidden Markov modelling techniques. We apply our multisite model to estimate the pandemic scope using publicly available disease count data including detected cases, recoveries among detected cases, and total deaths. These counts are used to estimate the case detection probability, the infection fatality rate through time, the probability of recovery, and several important population parameters including the rate of spread and importation of external cases. We estimate the total number of active COVID-19 cases per region of Canada for each reporting interval. We applied this multisite model Canada-wide to all provinces and territories, providing an estimate of the total COVID-19 burden for the 90 weeks from 23 April 2020 to 10 February 2022. We also applied this model to the five health authority regions of British Columbia, Canada, describing the pandemic in B.C. over the 31 weeks from 2 April 2020 to 30 October 2020.

We study the problem of data-driven background estimation, arising in the search of physics signals predicted by the Standard Model at the Large Hadron Collider. Our work is motivated by the search for the production of pairs of Higgs bosons decaying into four bottom quarks. A number of other physical processes, known as background, also share the same final state. The data arising in this problem is, therefore, a mixture of unlabeled background and signal events, and the primary aim of the analysis is to determine whether the proportion of unlabeled signal events is nonzero. A challenging but necessary first step is to estimate the distribution of background events. Past work in this area has determined regions of the space of collider events, where signal is unlikely to appear and where the background distribution is, therefore, identifiable. The background distribution can be estimated in these regions and extrapolated into the region of primary interest using transfer learning with a multivariate classifier. We build upon this existing approach in two ways. First, we revisit this method by developing a customized residual neural network which is tailored to the structure and symmetries of collider data. Second, we develop a new method for background estimation, based on the optimal transport problem, which relies on modeling assumptions distinct from earlier work. These two methods can serve as cross-checks for each other in particle physics analyses, due to the complementarity of their underlying assumptions. We compare their performance on simulated double Higgs boson data.

Reconstructing three-dimensional (3D) chromatin structure from conformation capture assays (such as Hi-C) is a critical task in computational biology, since chromatin spatial architecture plays a vital role in numerous cellular processes and direct imaging is challenging. Most existing algorithms that operate on Hi-C contact matrices produce reconstructed 3D configurations in the form of a polygonal chain. However, none of the methods exploit the fact that the target solution is a (smooth) curve in 3D: this contiguity attribute is either ignored or indirectly addressed by imposing spatial constraints that are challenging to formulate. In this paper we develop both B-spline and smoothing spline techniques for directly capturing this potentially complex 1D curve. We subsequently combine these techniques with a Poisson model for contact counts and compare their performance on a real data example. In addition, motivated by the sparsity of Hi-C contact data, especially when obtained from single-cell assays, we appreciably extend the class of distributions used to model contact counts. We build a general distribution-based metric scaling (DBMS) framework from which we develop zero-inflated and Hurdle Poisson models as well as negative binomial applications. Illustrative applications make recourse to bulk Hi-C data from IMR90 cells and single-cell Hi-C data from mouse embryonic stem cells.

Most businesses impose a supervisory hierarchy on employees to facilitate management, decision-making, and collaboration, yet routine inter-employee communication patterns within workplaces tend to emerge more naturally as a consequence of both supervisory relationships and the needs of the organization. What then is the relationship between a formal organizational structure and the emergent communications between its employees? Understanding the nature of this relationship is critical for the successful management of an organization. While scholars of organizational management have proposed theories relating organizational trees to communication dynamics, and separately, network scientists have studied the topological structure of communication patterns in different types of organizations; existing empirical analyses are both lacking in representativeness and limited in size. In fact, much of the methodology used to study the relationship between organizational hierarchy and communication patterns (and much of what is known about this relationship) comes from analyses of the Enron email corpus, reflecting a uniquely dysfunctional corporate environment. In this paper we develop new methodology for assessing the relationship between organizational hierarchy and communication dynamics and apply it to Microsoft Corporation, currently the highest valued company in the world, consisting of approximately 200,000 employees divided into 88 teams, organizational trees rooted at the senior leadership level. This reveals distinct communication network structures within and between teams. We then characterize the relationship of routine employee communication patterns to these team supervisory hierarchies, while empirically evaluating several theories of organizational management and performance. To do so, we propose new measures of communication reciprocity and new shortest-path distances for trees to track the frequency of messages passed up, down, and across the organizational hierarchy. By describing how communication clusters around the formal organization, we reveal the emergent communication dynamics between employees and the crucial role of position in the hierarchy.

We define a model for the joint distribution of multiple continuous la- tent variables, which includes a model for how their correlations depend on explanatory variables. This is motivated by and applied to social scientific re- search questions in the analysis of intergenerational help and support within families, where the correlations describe reciprocity of help between genera- tions and complementarity of different kinds of help. We propose an MCMC procedure for estimating the model which maintains the positive definiteness of the implied correlation matrices and describe theoretical results which jus- tify this approach and facilitate efficient implementation of it. The model is applied to data from the UK Household Longitudinal Study to analyse ex- changes of practical and financial support between adult individuals and their noncoresident parents.

Statistical modelling of individual-level event history data arising from varying observation schemes is a challenging problem, particularly due to unobserved and possibly misclassified individual states. Commonly used approaches rely on the hidden Markov models (HMM) to incorporate true underlying states. Each approach needs to account for the underlying data generating process and related external information and requires assumptions for estimation. This article develops a Bayesian HMM for natural history of colorectal cancer (CRC), combining data on latent disease states from randomised screening study and on observed clinical cancers from the population-based cancer registry. With our modelling approach and study design, we are able to provide estimates for latent state occupancy probabilities not only for screening-attenders but also for the control group and those who never attended screening—despite data on latent states only existing for the attenders. We use simulation-based calibration to ensure that posterior distributions can be reliably estimated despite the challenges brought in by the sampling scheme. We apply Bayesian computation to obtain posterior estimates of the quantities of interest. Two algorithms, Hamiltonian Monte Carlo (HMC) and Automatic Differentiation Variational Inference (ADVI), are applied and compared, first by using simulated data and then with a real data set. The modelling workflow can be applied for different cancer screening programmes and datasets which typically have similar challenges.

Marine mammals are increasingly vulnerable to human disturbance and climate change. Their diving behavior leads to limited visual access during data collection, making studying the abundance and distribution of marine mammals challenging. In theory, using data from more than one observation modality should lead to better informed predictions of abundance and distribution. With focus on North Atlantic right whales, we consider the fusion of two data sources to inform about their abundance and distribution. The first source is aerial distance sampling, which provides the spatial locations of whales detected in the region. The second source is passive acoustic monitoring (PAM), returning calls received at hydrophones placed on the ocean floor. Due to limited time on the surface and detection limitations arising from sampling effort, aerial distance sampling only provides a partial realization of locations. With PAM we never observe numbers or locations of individuals. To address these challenges, we develop a novel thinned point pattern data fusion. Our approach leads to improved inference regarding abundance and distribution of North Atlantic right whales throughout Cape Cod Bay, Massachusetts in the U.S. We demonstrate performance gains of our approach compared to that from a single source through both simulation and real data.

Despite its importance for insurance, there is almost no literature on statistical hail damage modeling. Statistical models for hailstorms exist, though they are generally not open-source, but no study appears to have developed a stochastic hail impact function. In this paper we use hail-related insurance claim data to build a Gaussian line process with extreme marks in order to model both the geographical footprint of a hailstorm and the damage to buildings that hailstones can cause. We build a model for the claim counts and claim values, and compare it to the use of a benchmark deterministic hail impact function. Our model proves to be better than the benchmark at capturing hail spatial patterns and allows for localized and extreme damage, which is seen in the insurance data. The evaluation of both the claim counts and value predictions shows that performance is improved compared to the benchmark, especially for extreme damage. Our model appears to be the first to provide realistic estimates for hail damage to individual buildings.

Injuries to the lower extremity joints are often debilitating, particularly for professional athletes. Understanding the onset of stressful conditions on these joints is, therefore, important in order to ensure prevention of injuries as well as individualised training for enhanced athletic performance. We study the biomechanical joint angles from the hip, knee and ankle for runners who are experiencing fatigue. The data is cyclic in nature and densely collected by body-worn sensors, which makes it ideal to work with in the functional data analysis (FDA) framework.

We develop a new method for multiple change point detection for functional data, which improves the state of the art with respect to at least two novel aspects. First, the curves are compared with respect to their maximum absolute deviation, which leads to a better interpretation of local changes in the functional data compared to classical ${\mathit{L}}^2$-approaches. Second, as slight aberrations are to be often expected in a human movement data, our method will not detect arbitrarily small changes but hunts for relevant changes, where maximum absolute deviation between the curves exceeds a specified threshold, say $\mathrm{\Delta }>0$. We recover multiple changes in a long functional time series of biomechanical knee angle data, which are larger than the desired threshold Δ, allowing us to identify changes purely due to fatigue. In this work we analyse data from both controlled indoor as well as from an uncontrolled outdoor (marathon) setting.

Monitoring key elements of disease dynamics (e.g., prevalence, case counts) is of great importance in infectious disease prevention and control, as emphasized during the COVID-19 pandemic. To facilitate this effort, we propose a new capture–recapture (CRC) analysis strategy that adjusts for misclassification stemming from the use of easily administered but imperfect diagnostic test kits, such as rapid antigen test-kits or saliva tests. Our method is based on a recently proposed “anchor stream” design, whereby an existing voluntary surveillance data stream is augmented by a smaller and judiciously drawn random sample. It incorporates manufacturer-specified sensitivity and specificity parameters to account for imperfect diagnostic results in one or both data streams. For inference to accompany case count estimation, we improve upon traditional Wald-type confidence intervals by developing an adapted Bayesian credible interval for the CRC estimator that yields favorable frequentist coverage properties. When feasible, the proposed design and analytic strategy provides a more efficient solution than traditional CRC methods or random sampling-based bias-corrected estimation to monitor disease prevalence while accounting for misclassification. We demonstrate the benefits of this approach through simulation studies and a numerical example that underscore its potential utility in practice for economical disease monitoring among a registered closed population.

In an effort to quantify and combat sexual assault, U.S. colleges and universities are required to disclose the number of reported sexual assaults on their campuses each year. However, many instances of sexual assault are never reported to authorities, and consequently, the number of reported assaults does not fully reflect the true total number of assaults that occurred; the reported values could arise from many combinations of reporting rate and true incidence. In this paper we estimate these underlying quantities via a hierarchical Bayesian model of the reported number of assaults. We use informative priors, based on national crime statistics, to act as a tiebreaker to help distinguish between reporting rates and incidence. We outline a Hamiltonian Monte Carlo (HMC) sampling scheme for posterior inference regarding reporting rates and assault incidence at each school and apply this method to campus sexual assault data from 2014–2019. Results suggest an increasing trend in reporting rates for the overall college population during this time. However, the extent of underreporting varies widely across schools. That variation has implications for how individual schools should interpret their reported crime statistics.

Mental health diseases which affect children’s lives and well-beings have received increased attention since the COVID-19 pandemic. Analyzing psychiatric clinical notes with topic models is critical to evaluating children’s mental status over time. However, few topic models are built for longitudinal settings, and most existing approaches fail to capture temporal trajectories for each document. To address these challenges, we develop a dynamic topic model with consistent topics and individualized temporal dependencies on the evolving document metadata. Our model preserves the semantic meaning of discovered topics over time and incorporates heterogeneity among documents. In particular, when documents can be categorized, we propose a classifier-free approach to maximize topic heterogeneity across different document groups. We also present an efficient variational optimization procedure adapted for the multistage longitudinal setting. In this case study, we apply our method to the psychiatric clinical notes from a large tertiary pediatric hospital in Southern California and achieve a 38% increase in the overall coherence of extracted topics. Our real data analysis reveals that children tend to express more negative emotions during state shutdowns and more positive when schools reopen. Furthermore, it suggests that sexual and gender minority (SGM) children display more pronounced reactions to major COVID-19 events and a greater sensitivity to vaccine-related news than non-SGM children. This study examines children’s mental health progression during the pandemic and offers clinicians valuable insights to recognize disparities in children’s mental health related to their sexual and gender identities.

Renewable energy is critical for combating climate change, whose first step is the storage of electricity generated from renewable energy sources. Li-ion batteries are a popular kind of storage units. Their continuous usage through charge-discharge cycles eventually leads to degradation. This can be visualized by plotting voltage discharge curves (VDCs) over discharge cycles. Studies of battery degradation have mostly concentrated on modeling degradation through one scalar measurement summarizing each VDC. Such simplification of curves can lead to inaccurate predictive models. Here we analyze the degradation of rechargeable Li-ion batteries from a NASA data set through modeling and predicting their full VDCs. With techniques from longitudinal and functional data analysis, we propose a new two-step predictive modeling procedure for functional responses residing on heterogeneous domains. We first predict the shapes and domain end points of VDCs using functional regression models. Then we integrate these predictions to perform a degradation analysis. Our functional approach allows the incorporation of usage information, produces predictions in a curve form and thus provides flexibility in the assessment of battery degradation. Through extensive simulation studies and cross-validated data analysis, our approach demonstrates better prediction than the existing approach of modeling degradation directly with aggregated data.

For decades National Football League (NFL) coaches’ observed fourth down decisions have been largely inconsistent with prescriptions based on statistical models. In this paper we develop a framework to explain this discrepancy using an inverse optimization approach. We model the fourth down decision and the subsequent sequence of plays in a game as a Markov decision process (MDP), the dynamics of which we estimate from NFL play-by-play data from the 2014 through 2022 seasons. We assume that coaches’ observed decisions are optimal but that the risk preferences governing their decisions are unknown. This yields an inverse decision problem for which the optimality criterion, or risk measure, of the MDP is the estimand. Using the quantile function to parameterize risk, we estimate which quantile-optimal policy yields the coaches’ observed decisions as minimally suboptimal. In general, we find that coaches’ fourth-down behavior is consistent with optimizing low quantiles of the next-state value distribution, which corresponds to conservative risk preferences. We also find that coaches exhibit higher risk tolerances when making decisions in the opponent’s half of the field, as opposed to their own half, and that league average fourth down risk tolerances have increased over time.

The paper targets the estimation of a poverty rate at the upazila level in Bangladesh through the use of demographic and health survey (DHS) data. Upazilas are administrative regions equivalent to counties or boroughs whose sample sizes are not large enough to provide reliable estimates or are even absent. We tackle this issue by proposing a small-area estimation model complementing survey data with remote sensing information at the area level. We specify an extended Beta mixed regression model within the Bayesian framework, allowing it to accommodate the peculiarities of sample data and to predict out-of-sample rates. Specifically, it enables to include estimates equal to either 0 or 1 and to model the strong intra-cluster correlation. We aim at proposing a method that can be implemented by statistical offices as a routine. In this spirit we consider a regularizing prior for coefficients, rather than a model selection approach, to deal with a large number of auxiliary variables. We compare our methods with existing alternatives using a design-based simulation exercise and illustrate its potential with the motivating application.

It is quite common to encounter compositional data in a regression framework in data analysis. When both responses and predictors are compositional, most existing models rely on a family of log-ratio based transformations to move the analysis from the simplex to the reals. This often makes the interpretation of the model more complex. A transformation-free regression model was recently developed, but it only allows for a single compositional predictor. However, many datasets include multiple compositional predictors of interest. Motivated by an application to hydrothermal liquefaction (HTL) data, a novel extension of this transformation-free regression model is provided that allows for two (or more) compositional predictors to be used via a latent variable mixture. A modified expectation-maximization algorithm is proposed to estimate model parameters, which are shown to have natural interpretations. Conformal inference is used to obtain prediction limits on the compositional response. The resulting methodology is applied to the HTL dataset. Extensions to multiple predictors are discussed.

Integrative analysis of multilevel pharmacogenomic data for modeling dependencies across various biological domains is crucial for developing genomic-testing based treatments. Chain graphs characterize conditional dependence structures of such multilevel data where variables are naturally partitioned into multiple ordered layers, consisting of both directed and undirected edges. Existing literature mostly focus on Gaussian chain graphs, which are ill-suited for nonnormal distributions with heavy-tailed marginals, potentially leading to inaccurate inferences. We propose a Bayesian robust chain graph model (RCGM) based on random transformations of marginals using Gaussian scale mixtures to account for node-level nonnormality in continuous multivariate data. This flexible modeling strategy facilitates identification of conditional sign dependencies among nonnormal nodes while still being able to infer conditional dependencies among normal nodes. In simulations we demonstrate that RCGM outperforms existing Gaussian chain graph inference methods in data generated from various nonnormal mechanisms. We apply our method to genomic, transcriptomic and proteomic data to understand underlying biological processes holistically for drug response and resistance in lung cancer cell lines. Our analysis reveals inter- and intra-platform dependencies of key signaling pathways to monotherapies of icotinib, erlotinib and osimertinib among other drugs, along with shared patterns of molecular mechanisms behind drug actions.

We propose a Bayesian meta-analysis to infer the current expansion rate of the Universe, called the Hubble constant (${\mathit{H}}_0$), via time delay cosmography. Inputs of the meta-analysis are estimates of two properties for each pair of gravitationally lensed images; time delay and Fermat potential difference estimates with their standard errors. A meta-analysis can be appealing in practice because obtaining each estimate from even a single lens system involves substantial human efforts, and thus estimates are often separately obtained and published. Moreover, numerous estimates are expected to be available once the Rubin Observatory starts monitoring thousands of strong gravitational lens systems. This work focuses on combining these estimates from independent studies to infer ${\mathit{H}}_0$ in a robust manner. The robustness is crucial because currently up to eight lens systems are used to infer ${\mathit{H}}_0$, and thus any biased input can severely affect the resulting ${\mathit{H}}_0$ estimate. For this purpose we adopt Student’s t error for the input estimates. We investigate properties of the resulting ${\mathit{H}}_0$ estimate via two simulation studies with realistic imaging data. It turns out that the meta-analysis can infer ${\mathit{H}}_0$ with subpercent bias and about 1% level of coefficient of variation, even when 30% of inputs are manipulated to be outliers. We also apply the meta-analysis to three gravitationally lensed systems to obtain an ${\mathit{H}}_0$ estimate and compare it with existing estimates. An R package $\mathtt{h}\mathtt{0}$ is publicly available for fitting the proposed meta-analysis.

When using electronic health records (EHRs) for clinical and translational research, additional data is often available from external sources to enrich the information extracted from EHRs. For example, academic biobanks have more granular data available, and patient reported data is often collected through small-scale surveys. It is common that the external data is available only for a small subset of patients who have EHR information. We propose efficient and robust methods for building and evaluating models for predicting the risk of binary outcomes using such integrated EHR data. Our method is built upon an idea derived from the two-phase design literature that modeling the availability of a patient’s external data as a function of an EHR-based preliminary predictive score leads to effective utilization of the EHR data. Through both theoretical and simulation studies, we show that our method has high efficiency for estimating log-odds ratio parameters, the area under the ROC curve, as well as other measures for quantifying predictive accuracy. We apply our method to develop a model for predicting the short-term mortality risk of oncology patients, where the data was extracted from the University of Pennsylvania hospital system EHR and combined with survey-based patient reported outcome data.

In this paper we study the number of inpatient admissions by individuals to hospital emergency rooms reported by the 2003 Medical Expenditure Panel Survey (MEPS), which the United States Agency for Health Research and Quality conducts. Explanatory variables such as health status, access, use, and costs of health services in the U.S.A. are considered. Our main goal is to properly model the number of inpatient admissions, according to the geographical U.S. regions, as a tool for measuring the volume of diagnostic procedures in the health care system. In the analysis four clusters were determined according to the regions in the U.S., namely, the midwest, northeast, south, and west. The clustered analysis of this count data from the MEPS is a novel contribution to the best of our knowledge. Our analysis demonstrated that a clustered negative binomial (CNB) regression (Poisson model with latent gamma effects) might not be a suitable choice for analyzing the MEPS data. This fact motivates us to introduce a new regression model to handle clustered count data. To account for correlation within clusters, we propose a Poisson regression model where the observations within the same cluster are driven by the same latent random effect that follows a Birnbaum–Saunders distribution with a parameter that controls the strength of dependence among the individuals. This novel multivariate count model is called Clustered Poisson–Birnbaum–Saunders (CPBS) regression. The CPBS model is analytically tractable, and its moment structure can be explicitly obtained. We also derive theoretical/methodological studies to advise when the Birnbaum–Saunders effect should be preferred over the gamma effect (and vice-versa) in terms of probability tail. Estimation is performed through the maximum likelihood method. Here we also developed an expectation-maximization (EM) algorithm for estimation. Simulation results that evaluate the finite-sample performance of our proposed estimators are presented. Studies on the potential impact of model misspecification were conducted, and comparisons between our model and a CNB regression were also addressed. A full statistical analysis of the MEPS data reveals that, compared to the CNB model, the CPBS regression model produces better results in terms of prediction and goodness-of-fit.

Analyses of occurrences of residential burglary in urban areas have shown that crime rates are not spatially homogeneous: rates vary across the network of city streets, resulting in some areas being far more susceptible to crime than others. The explanation for why a certain segment of the city experiences high crime may be different than why a neighboring area experiences high crime. Motivated by the importance of understanding spatial patterns such as these, we consider a statistical model of burglary defined on the street network of Boston, Massachusetts. Leveraging ideas from functional data analysis, our proposed solution consists of a generalized linear model with vertex-indexed covariates, allowing for an interpretation of the covariate effects at the street level. We employ a regularization procedure cast as a prior distribution on the regression coefficients under a Bayesian setup so that the predicted responses vary smoothly according to the connectivity of the city. We introduce a novel variable selection procedure, examine computationally efficient methods for sampling from the posterior distribution of the model parameters, and demonstrate the flexibility of our proposed modeling structure. The resulting model and interpretations provide insight into the spatial network patterns and dynamics of residential burglary in Boston.

This paper proposes a novel methodology called the mixture of Bayesian predictive syntheses (MBPS) for multiple time series count data for the challenging task of predicting the numbers of COVID-19 inpatients and isolated cases in Japan and Korea at the subnational level. MBPS combines a set of predictive models and partitions the multiple time series into clusters based on their contribution to predicting the outcome. In this way MBPS leverages the shared information within each cluster and is suitable for predicting COVID-19 inpatients since the data exhibit similar dynamics over multiple areas. Also, MBPS avoids using a multivariate count model, which is generally cumbersome to develop and implement. Our Japanese and Korean data analyses demonstrate that the proposed MBPS methodology has improved predictive accuracy and uncertainty quantification.

Dynamic networks have been increasingly used to characterize brain connectivity that varies during resting and task states. In such characterizations a connectivity network is typically measured at each time point for a subject over a common set of nodes representing brain regions, together with rich subject-level information. A common approach to analyzing such data is an edge-based method that models the connectivity between each pair of nodes separately. However, such approach may have limited performance when the noise level is high and the number of subjects is limited, as it does not take advantage of the inherent network structure. To better understand if and how the subject-level covariates affect the dynamic brain connectivity, we introduce a semiparametric dynamic network response regression that relates a dynamic brain connectivity network to a vector of subject-level covariates. A key advantage of our method is to exploit the structure of dynamic imaging coefficients in the form of high-order tensors. We develop an efficient estimation algorithm and evaluate the efficacy of our approach through simulation studies. Finally, we present our results on the analysis of a task-related study on social cognition in the Human Connectome Project, where we identify known sex-specific effects on brain connectivity that cannot be inferred using alternative methods.

We are motivated by a study that seeks to better understand the dynamic relationship between muscle activation and paw position during locomotion. For each gait cycle in this experiment, activation in the biceps and triceps is measured continuously and in parallel with paw position as a mouse trotted on a treadmill. We propose an innovative general regression method that draws from both ordinary differential equations and functional data analysis to model the relationship between these functional inputs and responses as a dynamical system that evolves over time. Specifically, our model addresses gaps in both literatures and borrows strength across curves estimating ODE parameters across all curves simultaneously rather than separately modeling each functional observation. Our approach compares favorably to related functional data methods in simulations and in cross-validated predictive accuracy of paw position in the gait data. In the analysis of the gait cycles, we find that paw speed and position are dynamically influenced by inputs from the biceps and triceps muscles and that the effect of muscle activation persists beyond the activation itself.

We model longitudinal macular thickness measurements to monitor the course of glaucoma and prevent vision loss due to disease progression. The macular thickness varies over a 6 × 6 grid of locations on the retina, with additional variability arising from the imaging process at each visit. Currently, ophthalmologists estimate slopes using repeated simple linear regression for each subject and location. To estimate slopes more precisely, we develop a novel Bayesian hierarchical model for multiple subjects with spatially varying population-level and subject-level coefficients, borrowing information over subjects and measurement locations. We augment the model with visit effects to account for observed spatially correlated visit-specific errors. We model spatially varying: (a) intercepts, (b) slopes, and (c) log-residual standard deviations (SD) with multivariate Gaussian process priors with Matérn cross-covariance functions. Each marginal process assumes an exponential kernel with its own SD and spatial correlation matrix. We develop our models for and apply them to data from the Advanced Glaucoma Progression Study. We show that including visit effects in the model reduces error in predicting future thickness measurements and greatly improves model fit.

We propose a hierarchical Bayesian model to infer RNA synthesis, processing, and degradation rates from time-course RNA sequencing data, based on an ordinary differential equation system that models the RNA life cycle. We parametrize the latent kinetic rates, which rule the system, with a novel functional form and estimate their parameters through three Dirichlet process mixture models. Owing to the complexity of this approach, we are able to simultaneously perform inference, clustering, and model selection. We apply our method to investigate transcriptional and post-transcriptional responses of murine fibroblasts to the activation of the proto-oncogene Myc. Our approach uncovers simultaneous regulations of the rates, which had been largely missed in previous analyses of this biological system.

High-dimensional spectral data—routinely generated in dairy production—are used to predict a range of traits in milk products. Partial least squares (PLS) regression is ubiquitously used for these prediction tasks. However, PLS regression is not typically viewed as arising from a probabilistic model, and parameter uncertainty is rarely quantified. Additionally, PLS regression does not easily lend itself to model-based modifications, coherent prediction intervals are not readily available, and the process of choosing the latent-space dimension, $\mathtt{Q}$, can be subjective and sensitive to data size.

We introduce a Bayesian latent-variable model, emulating the desirable properties of PLS regression while accounting for parameter uncertainty in prediction. The need to choose $\mathtt{Q}$ is eschewed through a nonparametric shrinkage prior. The flexibility of the proposed Bayesian partial least squares (BPLS) regression framework is exemplified by considering sparsity modifications and allowing for multivariate response prediction.

The BPLS regression framework is used in two motivating settings: (1) multivariate trait prediction from mid-infrared spectral analyses of milk samples and (2) milk pH prediction from surface-enhanced Raman spectral data. The prediction performance of BPLS regression at least matches that of PLS regression. Additionally, the provision of correctly calibrated prediction intervals objectively provides richer, more informative inference for stakeholders in dairy production.

Do the impacts that occur when playing high school football have concussive effects that accelerate cognitive decline late in life? We examine this possibility using newly available cognitive data describing people in 2020 who graduated high school in 1957. Someone who was 18 in 1957 would be 81 in 2020. For this comparison we develop a new design for an observational study, called a triples design, and discuss its advantages and construction. A triples design consists of M blocks of size 3, where a block contains either one treated individual and two controls or two treated individuals and one control. A triples design is the simplest design that uses weights, with just two weights. The “entire number” is $\{1-\mathit{e}(\mathbf{x})\}/\mathit{e}(\mathbf{x})$, where $\mathit{e}(\mathbf{x})$ is the propensity score at covariate x, so it is the ratio of controls-to-treated expected at x. Unlike a matched pairs design, which can remove the bias from observed covariates when the “entire number” exceeds 1, the triples design can succeed when the entire number exceeds 1/2, reflecting the possibility of matching two treated individuals to the same control. Like full matching, a triples design can match more people than can matched pairs, yet have smaller within-block covariate distances. Unlike full matching, there are no matched pairs. Like matching with multiple controls, a triples design will have a larger design sensitivity than a design which includes matched pairs, under simple models for continuous outcomes; that is, in favorable situations the design is expected to report greater insensitivity to unmeasured biases. Because there are just two weights, it is easy to construct weighted graphics for exploratory displays from triples designs. A heuristic algorithm containing network optimization constructs the design.

Mutations in the noncoding DNA, which represents approximately 99% of the human genome, have been crucial to understanding disease mechanisms through dysregulation of disease-associated genes. One key element in gene regulation that noncoding mutations mediate is the binding of proteins to DNA sequences. Insertion and deletion of bases (InDels) are the second most common type of mutations, following single nucleotide polymorphisms, that may impact protein-DNA binding. However, no existing methods can estimate and test the effects of InDels on the process of protein-DNA binding. We develop a novel test of statistical significance, namely, the binding change test (BC test), using a Markov model to evaluate the impact and identify InDels altering protein-DNA binding. The test predicts binding changer InDels of regulatory significance with an efficient importance sampling algorithm generating background sequences in favor of large binding affinity changes. Simulation studies demonstrate its excellent performance. The application to human leukemia data uncovers, in critical cis-regulatory elements, candidate pathological InDels on modulating TF binding in leukemic patients. We develop an R package atIndel, which is available on GitHub.

Cybersecurity is an important issue given the increasing risks due to cyberattacks in many areas. Cyberattacks could result in huge losses such as data breaches, failures in the control systems of infrastructures, physical damages in manufacturing industries, etc. As a result, cybersecurity-related research has grown rapidly for in-depth analysis. One main interest is to understand the correlated nature of cyberattack data. To understand such characteristics, we propose a spatio-temporal model for the hostwisely aggregated cyberattack data by incorporating the characteristics of the attackers. We develop a new dissimilarity measure as a proxy of spatial distance to be integrated into the model. The proposed model can be considered as a spatial extension of the GARCH model. The estimation is carried out using a Bayesian approach, which is demonstrated to work well in simulations. The proposed model is applied to publicly available honeypot data after the data are divided by selected features of the attackers via clustering. The estimated model parameters vary by groups of attackers, which was not revealed by modeling the entire dataset.

Understanding cell type composition and gene expression of spatial transcriptomic data is crucial for comprehending phenotypic variability and detecting key factors that influence disease susceptibility of complex traits. Detecting cell type specific expression patterns from spatial transcriptome profiles is important in studying the cellular components and gene expression of individual cell classes and structural architecture. In this paper we develop mixed effect multiplicative-additive Poisson-gamma models to analyze spatial (MAPS) transcriptomics data using cell type-specific gene expressions in single cell RNA-sequenceing (scRNA-seq) data. To build the mixed effect multiplicative-additive Poisson-gamma models, the gene expression counts of spatial transcriptomics data are treated as dependent variables, and the mean and variance parameters of scRNA-seq data are used to construct independent variables to explain the dependent variables on the basis of Poisson-gamma mixture. One novelty of the proposed mixed models is that the variance parameters of scRNA-seq are used to describe the within-cell-type variations or stochasticity. We develop iteratively analytical formulae to estimate the cell type proportions and dispersion parameters. To address the important research problems and help with intensive spatial transcriptomics data analysis, a readily available software, MAPS, is developed to implement the proposed methods. By simulation study and real data analysis, MAPS is found to perform better than or similar to robust cell type decomposition (RCTD), SpatialDWLS (dampened weighted least squares), conditional autoregressive-based deconvolution (CARD), and a Spatially weighted pOissoN-gAmma Regression model (SONAR). Computationally, MAPS is significantly faster than RCTD and SpatialDWLS. MAPS provides a novel way for mapping spatial tissue architecture.

Effective integration of single-cell data can facilitate the discovery of cell-type specific gene expression patterns and cellular interactions, ultimately leading to a better understanding of various biological processes and diseases. However, datasets from different platforms, species, and modalities exhibit various levels of heterogeneities, posing significant challenges in data alignment using a unified approach. Here we propose DeepMap, a flexible and efficient method for single-cell data integration, by taking advantage of the deep learning framework. Our method utilizes iterative cell matching based on mutual nearest neighbors, leverages an autoencoder framework to learn harmonized representations of cells from various datasets, and incorporates a covariance penalty term into the framework for structure preservation. In addition to harmonization of data from different datasets, we specifically take account of the preservation of important biological variations within dataset, which is crucial to reliable downstream analysis. Comprehensive real data analysis demonstrates the flexibility of DeepMap for diverse datasets from different platforms, species, and modalities, and highlights its marked ability in preserving structures over existing integration methods with enhanced computational efficiency and optimized memory usage. The robust DeepMap-integrated data offers promising prospects for advancing our understanding of cell biology, hence making it a highly attractive option for integrative single-cell data analysis.