Bayesian State-space Model Research Articles

Local effects of non-pharmaceutical interventions on mitigation of COVID-19 spread through decreased human mobilities in Japan: a prefecture-level mediation analysis

To control the COVID-19 epidemic, the Japanese government and the local governments have repeatedly implemented non-pharmaceutical interventions (NPIs) throughout 2020–2022. Using Bayesian state-space mediation models, we examined the effect of repeated NPIs on infection spread mitigation, mediated by human mobility changes in each prefecture during three epidemic phases: from April 1, 2020 to February 28, 2021; from March 1, 2021 to December 16, 2021; and from December 17, 2021 to December 31, 2022. In the first phase, controlling downtown populations at nighttime was effective in mitigating the infection spread in almost all prefectures. In the second and third phases, the effect was not clear, especially in metropolitan prefectures. Controlling visitors from the central prefectures of metropolitan areas was effective in mitigating infection spread in the surrounding prefectures during all phases. These results suggest that the local spread of infection can be mitigated by focusing on nighttime human mobility control in downtown areas before the epidemic spreads widely and transmission routes become more diverse, and that the geospatial spread of infection can be prevented by controlling the flows of people from large cities to other areas.

Bayesian State-space modelling of stock Markets in G7 Countries

This work examines the impact of Coronavirus disease (COVID-19) on the stock market of Group of Seven (G7) countries during the first wave of COVID-19 using daily data from March, 1st of 2020 to December, 31st of 2020. Focussing on such period, a Bayesian Structural Time Series Model (BSTSM) was used to capture the effects of first wave of COVID-19 on the stock market performance of these G7 countries by employing a Markov Chain Monte Carlo (MCMC) method. We considered an Autoregressive (AR) model with time-varying parameters and a local linear trend model to know if the stock price of these countries during the period of the first wave of COVID-19 is changing over time. There was a stochastic trend in stock prices of G7 countries during the period of the first wave of COVID-19 while the AR process itself was also changing over time. The stock market of the USA followed by Japan performed better than other G7 countries during the first phase of the COVID-19 pandemic while the stock market of France was affected during the COVID-19 pandemic.

Variability in somatic growth over time and space determines optimal season-opening date in the Oregon ocean shrimp (Pandalus jordani) fishery

Growth variability is a key contributor to the dynamic productivity of populations. Thus, better understanding and accounting for variation in growth can improve both tactical and strategic management. Using ocean shrimp (Pandalus jordani) from the U.S. West Coast as a case study, we demonstrate interactions between growth and optimal fishery opening dates. While the fishery opens on April 1, industry often delays fishing to minimize catches of small shrimp. Understanding drivers of size-at-recruitment can help managers optimize opening dates and shrimpers plan their participation in this and other fisheries. Using three decades of fishery-dependent sampling, we built a spatially-, temporally-, and environmentally-explicit Bayesian state-space model for shrimp size-at-age. Model outputs were then used to parameterize a revenue-per-recruit model and explore how variability in size-at-recruitment impacted optimal opening dates. Shrimp recruited at smaller sizes farther north. Delaying opening would likely benefit shrimpers in areas and years with smaller shrimp and higher fishing mortality. Broadly, choosing when to open the fishery is a complex decision requiring understanding of growth, but also recruitment, economic incentives, and life history.

The African Penguin Spheniscus demersus should be considered Critically Endangered

The African Penguin Spheniscus demersus has been considered a threatened species since 1984 and, aside from a short period around 2000, its population has been in constant decline since at least the 1950s. By combining counts of the numbers of breeding pairs made at 26 colonies in South Africa and Namibia between 1979 and 2023 with Bayesian state–space models we reassess the species’ conservation status. The breeding population has declined by 77.8% (95% credible intervals: 71.8–84.6%) over the last 30 years (3 generations), from ∼44 300 breeding pairs in 1993 to ∼9 900 pairs in 2023. This falls just below the threshold for a global IUCN Red List status of Critically Endangered (CR) under criterion A2. However, the decline in Namibia exceeds that threshold at the national level (30-year decline = 82.4%, 78.2–86.2%) following the loss of ∼3 600 breeding pairs there in the last 5 years. In South Africa, the Western Cape population is now declining at <1.5% per annum, but the annual rate of decline in the Eastern Cape has worsened substantially, reaching 13% over the last 10 years. Overall, the global population has more than halved in the last decade and has fallen below 10 000 pairs for the first time. Moreover, “moving window reductions” over 3 generations using observed and projected population trajectories (up to the year 2033) indicate that the decline of the African Penguin population will exceed the 80% CR threshold (under criterion A4ab) with a high probability by 2028. Accordingly, we propose that the African Penguin should be transferred to the IUCN Red List category Critically Endangered.

Hierarchical Bayesian Integrated Modeling of Age- and Sex-Structured Wildlife Population Dynamics

AbstractBiodiversity of large wild mammals is declining at alarming rates worldwide. It is therefore imperative to develop effective population conservation and recovery strategies. Population dynamics models can provide insights into processes driving declines of particular populations of a species and their relative importance. But there are insufficient tools, namely population dynamics models for wild herbivores, for characterizing their decline and for guiding conservation and management actions. Therefore, we have developed a model which can serve as a tool to fill that void. Specifically, we develop an integrated Bayesian state-space population dynamics model for wildlife populations and illustrate it using a topi population inhabiting the Greater Mara-Serengeti Ecosystem in Kenya and Tanzania. The model integrates ground demographic survey with aerial survey monitoring data. It incorporates population age and sex structure and life history traits and strategies and relates birth rates, age-specific survival rates and sex ratios with meteorological covariates, prior population density, environmental seasonality and predation risk. It runs on a monthly time step, enabling accurate characterization of reproductive seasonality, phenology, synchrony and prolificacy of births, juvenile and adult recruitments. Model performance is evaluated using balanced bootstrap sampling and by comparing model predictions with empirical aerial population size estimates. The hierarchical Bayesian model is implemented using MCMC methods for parameter estimation, prediction and inference and reproduces several well-known features of the Mara topi population, including striking and persistent population decline, seasonality of births, juvenile and adult recruitments. It is general and can be readily adapted for other wildlife species and extended to incorporate several additional useful features. Supplementary materials accompanying this paper appear on-line.

Impact of COVID-19 pandemic on breast cancer screening in a large midwestern United States academic medical center.

Access to breast screening mammogram services decreased during the COVID-19 pandemic. Our objectives were to estimate: 1) the COVID-19 affected period, 2) the proportion of pandemic-associated missed or delayed screening encounters, and 3) pandemic-associated patient attrition in screening encounters overall and by sociodemographic subgroup. We included screening mammogram encounter EPIC data from 1-1-2019 to 12-31-2022 for females ≥40 years old. We used Bayesian State Space models to describe weekly screening mammogram counts, modeling an interruption that phased in and out between 3-1-2020 and 9-1-2020. We used the posterior predictive distribution to model differences between a predicted, uninterrupted process and the observed screening mammogram counts. We estimated associations between race/ethnicity and age group and return screening mammogram encounters during the pandemic among those with 2019 encounters using logistic regression. Our analysis modeling weekly screening mammogram counts included 231,385 encounters (n = 127,621 women). Model-estimated screening mammograms dropped by >98% between 03-15-2020 and 05-24-2020 followed by a return to pre-pandemic levels or higher with similar results by race/ethnicity and age group. Among 79,257 women, non-Hispanic (NH) Asians, NH Blacks, and Hispanics had significantly (p < .05) lower odds of screening encounter returns during 2020-2022 vs. NH Whites with odds ratios (ORs) from 0.70 to 0.91. Among 79,983 women, those 60-69 had significantly higher odds of any return screening encounter during 2020-2022 (OR = 1.28), while those ≥80 and 40-49 had significantly lower odds (ORs 0.77, 0.45) than those 50-59 years old. A sensitivity analysis suggested a possible pre-existing pattern. These data suggest a short-term pandemic effect on screening mammograms of ~2 months with no evidence of disparities. However, we observed racial/ethnic disparities in screening mammogram returns during the pandemic that may be at least partially pre-existing. These results may inform future pandemic planning and continued efforts to eliminate mammogram screening disparities.

Design of an Iterative Method for Enhanced Multimodal Time Series Analysis Using Graph Attention Networks, Variational Graph Autoencoders, and Transfer Learning

In the ever-evolving landscape of data analysis, the need to efficiently and accurately interpret multimodal time series data has become paramount. Traditional methods often fall short in addressing the complex dependencies and dynamics inherent in such data, limiting their effectiveness in real-world applications. This work introduces a comprehensive approach that leverages Graph Attention Networks (GATs), Variational Graph Autoencoders (VGAEs), transfer learning with pretrained transformers, and Bayesian state-space models to overcome these limitations. GATs are selected for their ability to dynamically focus on relevant modalities through attention mechanisms, thereby capturing the intricate relationships between different data modalities. This method significantly enhances the model's ability to integrate multimodal information, leading to notable improvements in classification, prediction, and anomaly detection tasks. VGAEs are utilized to learn latent representations within a graph-based framework, promoting unsupervised learning while unveiling the underlying data structure. The resultant embeddings are pivotal for downstream tasks like clustering and visualization, encapsulating the interactions within multimodal time series data effectively. Furthermore, this work incorporates transfer learning with pretrained transformers to harness extensive knowledge from large datasets, adapting it to multimodal time series analysis. This strategy excels in capturing long-range dependencies, thereby augmenting generalization and performance in data-scarce scenarios. Bayesian state-space models are employed to elucidate the temporal dynamics and uncertainties of time series data, offering a robust framework for probabilistic inference and enhancing the interpretability and reliability of forecasting and anomaly detection. The efficacy of the proposed model is rigorously evaluated using diverse datasets, including the Yahoo! Stock Dataset, Forest Cover Dataset, and an empirical collection of 100k time series data samples. The results demonstrate a significant leap in performance metrics, including a 9.5% increase in precision, 8.5% boost in accuracy, 8.3% rise in recall, 10.4% reduction in delay, 9.4% enhancement in AUC, and a 5.9% improvement in specificity, alongside superior pre-emption capabilities compared to existing methods. This work not only addresses the pressing need for advanced multimodal time series analysis techniques but also sets a new benchmark for efficiency and accuracy. The integration of GATs, VGAEs, transfer learning with pretrained transformers, and Bayesian state-space models presents a formidable approach that significantly advances the field, offering profound impacts on a wide array of applications.

Spectral dynamic causal modeling: A didactic introduction and its relationship with functional connectivity.

We present a didactic introduction to spectral dynamic causal modeling (DCM), a Bayesian state-space modeling approach used to infer effective connectivity from noninvasive neuroimaging data. Spectral DCM is currently the most widely applied DCM variant for resting-state functional MRI analysis. Our aim is to explain its technical foundations to an audience with limited expertise in state-space modeling and spectral data analysis. Particular attention will be paid to cross-spectral density, which is the most distinctive feature of spectral DCM and is closely related to functional connectivity, as measured by (zero-lag) Pearson correlations. In fact, the model parameters estimated by spectral DCM are those that best reproduce the cross-correlations between all measurements-at all time lags-including the zero-lag correlations that are usually interpreted as functional connectivity. We derive the functional connectivity matrix from the model equations and show how changing a single effective connectivity parameter can affect all pairwise correlations. To complicate matters, the pairs of brain regions showing the largest changes in functional connectivity do not necessarily coincide with those presenting the largest changes in effective connectivity. We discuss the implications and conclude with a comprehensive summary of the assumptions and limitations of spectral DCM.

Abstract 4797: Impact of COVID-19 pandemic on breast cancer screening and diagnosis stage in a large midwestern United States academic medical center

Abstract Background: Access to breast cancer screening mammogram services decreased in association with the COVID-19 pandemic. There is also evidence for increases in late-stage breast cancer diagnoses. Our objectives were to determine: 1) the COVID-19-affected period on mammogram screening, 2) the proportion of pandemic-associated missed or delayed mammogram screening visits overall and by race/ethnicity and age group, and 3) evidence for pandemic-associated shifts in diagnosis stage. Methods: Screening mammogram encounter data between 1-1-2019 and 12-31-2022 were extracted from EPIC for females ≥ 40 years old for the screening analysis. We used Bayesian state space models to describe weekly screening mammogram counts, modeling an interruption that phased in and out from 3-1-2020 to 9-1-2020. We used the posterior predictive distribution to simulate differences between a predicted, uninterrupted process and the observed screening mammogram counts. Breast cancer diagnoses at ≥ 21 years from the tumor registry between 12-1-2018 and 11-30-2021 were included in the stage analysis. We used logistic regression models to estimate late-stage diagnosis odds comparing matched three-month periods during the pandemic to before the pandemic. Results: A total of 319,492 encounters among 146,644 women were included. Model-estimated screening mammograms dropped by 98.8% (95% CI 95.1 to 100) between 3-15-2020 and 5-24-2020, returning to pre-pandemic levels or higher after this period. Drops in screening mammogram encounters did not vary significantly by race/ethnicity or age group (p &amp;gt; .75). Among 4,669 breast cancer diagnoses, we found no significant differences in the odds of late-stage diagnoses for any period vs. the same period before the pandemic (Table). Conclusions: These data suggest a short-term pandemic effect on screening mammograms. Evidence for increases in late-stage diagnoses is limited. These results may inform future pandemic planning. OR (95% CI) Period comparison 1.19 (0.74 to 1.91) 3/2020 to 5/2020 vs. 3/2019 to 5/2019 1.05 (0.69 to 1.60) 6/2020 to 8/2020 vs. 6/2019 to 8/2019 1.14 (0.73 to 1.78) 9/2020 to 11/2020 vs. 9/2019 to 11/2019 0.92 (0.60 to 1.41) 12/2020 to 2/2021 vs. 12/2019 to 2/2020 1.08 (0.71 to 1.65) 3/2021 to 5/2021 vs. 3/2019 to 5/2019 0.89 (0.59 to 1.33) 6/2021 to 8/2021 vs. 6/2019 to 8/2019 1.31 (0.84 to 2.04) 9/2021 to 11/2021 vs. 9/2019 to 11/2019 Citation Format: Kimberly J. Johnson, RJ Waken, Caitlin P. O'Connell, Derek Brown. Impact of COVID-19 pandemic on breast cancer screening and diagnosis stage in a large midwestern United States academic medical center [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 4797.

Bayesian state-space synthetic control method for deforestation baseline estimation for forest carbon credits

Abstract Carbon credits from the reducing emissions from deforestation and degradation (REDD+) projects have been criticized for issuing junk carbon credits due to invalid ex-ante baselines. Recently, the concept of ex-post baseline has been discussed to overcome the criticism, while ex-ante baseline is still necessary for project financing and risk assessment. To address this issue, we propose a Bayesian state-space model that integrates ex-ante baseline projection and ex-post dynamic baseline updating in a unified manner. Our approach provides a tool for appropriate risk assessment and performance evaluation of REDD+ projects. We apply the proposed model to a REDD+ project in Brazil and show that it may have had a small, positive effect but has been overcredited. We also demonstrate that the 90% predictive interval of the ex-ante baseline includes the ex-post baseline, implying that our ex-ante estimation can work effectively.

Density-dependent dynamics help explain the simultaneous expansion and decline of woodlands in the western US

Woodlands across the globe have expanded in recent centuries in response to shifts in climate, land use, and disturbance patterns, but are expected to be particularly vulnerable to decline as a result of environmental change. Gaps in our understanding of the drivers of woodland spatial dynamics, however, limit our ability to forecast range shifts. While the effect of extrinsic factors, such as fire and climate, on woodlands are well-documented, we lack knowledge on the role of intrinsic factors such as density-dependence in driving tree abundances and distributions across space and time. We used two remotely-sensed datasets of tree cover and climate predictors in Bayesian state-space models to test the effects of climate and density on the spatiotemporal changes in pinyon pine (Pinus monophylla) cover in the Great Basin, USA. We compared models with only climate predictors to models including both climate and tree cover (i.e. density-dependence). Models including density-dependence better explained observed dynamics than climate alone. Pinyon pine cover increased at the fastest rate in areas with intermediate precipitation and low maximum temperatures. Spatial effects of climate interacted with temporal variation in climate. In cool, wet areas, growth was higher in warm, dry periods. In warm, dry areas, it was favorable to have cool periods. However, positive direct effects of climate at mid to high elevation were partially offset by density-dependence. Because extant density was positively correlated with average precipitation and negatively correlated with temperature, pinyon pine cover often declined in climatically favorable areas. Our results indicate that direct effects of warming and drying climate will likely lead to increases in cover at high elevations and declines at low elevations in the long-term. However, the added effects of density-dependence may create counter-intuitive patterns with increases in low density, low elevation areas in the near term and declines in seemingly favorable high density cooler-wetter locations. As a result, management intended to increase woodland resilience through silvicultural treatments like density reductions may have the greatest impact at intermediate elevations where increasing climate stress coincides with high tree densities.

A point mass proposal method for Bayesian state-space model fitting

State-space models (SSMs) are commonly used to model time series data where the observations depend on an unobserved latent process. However, inference on the model parameters of an SSM can be challenging, especially when the likelihood of the data given the parameters is not available in closed-form. One approach is to jointly sample the latent states and model parameters via Markov chain Monte Carlo (MCMC) and/or sequential Monte Carlo approximation. These methods can be inefficient, mixing poorly when there are many highly correlated latent states or parameters, or when there is a high rate of sample impoverishment in the sequential Monte Carlo approximations. We propose a novel block proposal distribution for Metropolis-within-Gibbs sampling on the joint latent state and parameter space. The proposal distribution is informed by a deterministic hidden Markov model (HMM), defined such that the usual theoretical guarantees of MCMC algorithms apply. We discuss how the HMMs are constructed, the generality of the approach arising from the tuning parameters, and how these tuning parameters can be chosen efficiently in practice. We demonstrate that the proposed algorithm using HMM approximations provides an efficient alternative method for fitting state-space models, even for those that exhibit near-chaotic behavior.

Stock status of kawakawa Euthynnus affinis (Cantor, 1849) fishery using surplus production model: An assessment from coastal waters of Tamil Nadu, Bay of Bengal, Southeast coast of India

Kawakawa Euthynnus affinis (Cantor, 1849) is one of the commercially exploited tuna fishery resource of India. Except landing data, no other information was available on its stock status. Hence, present study investigated the sustainability status of kawakawa fishery from Tamil Nadu, India. Data on annual catch and effort were reconstructed for the period of 2001–2020. The Catch-based MSY (CMSY), the Bayesian state-space Schaefer surplus production Model (BSM), and stochastic Surplus Production model in Continuous Time (SPiCT) packages were used to investigate the stock exploitation by running surplus production models with catch, catch per unit effort (CPUE), and other life history data. The biological reference points result of maximum sustainable yield (MSY) or biomass giving MSY (Bmsy) and fishing mortality giving MSY (Fmsy) of all the models indicated that the kawakawa stock in Tamil Nadu was subjected to overfishing and heading towards overfished status in recent decades. Results of the present study suggest reduction of fishing effort to keep the landing at MSY level and achieve a sustainable fishery.

Bayesian multivariate longitudinal model for immune responses to Leishmania: A tick-borne co-infection study.

While many Bayesian state-space models for infectious disease processes focus on population infection dynamics (eg, compartmental models), in this work we examine the evolution of infection processes and the complexities of the immune responses within the host using these techniques. We present a joint Bayesian state-space model to better understand how the immune system contributes to the control of Leishmania infantum infections over the disease course. We use longitudinal molecular diagnostic and clinical data of a cohort of dogs to describe population progression rates and present evidence for important drivers of clinical disease. Among these results, we find evidence for the importance of co-infection in disease progression. We also show that as dogs progress through the infection, parasite load is influenced by their age, ectoparasiticide treatment status, and serology. Furthermore, we present evidence that pathogen load information from an earlier point in time influences its future value and that the size of this effect varies depending on the clinical stage of the dog. In addition to characterizing the processes driving disease progression, we predict individual and aggregate patterns of Canine Leishmaniasis progression. Both our findings and the application to individual-level predictions are of direct clinical relevance, presenting possible opportunities for application in veterinary practice and motivating lines of additional investigation to better understand and predict disease progression. Finally, as an important zoonotic human pathogen, these results may support future efforts to prevent and treat human Leishmaniosis.

Introducing shrinkage in heavy-tailed state space models to predict equity excess returns

AbstractWe forecast excess returns of the S &amp;P 500 index using a flexible Bayesian econometric state space model with non-Gaussian features at several levels. More precisely, we control for overparameterization via global–local shrinkage priors on the state innovation variances as well as the time-invariant part of the state space model. The shrinkage priors are complemented by heavy tailed state innovations that cater for potential large breaks in the latent states, even if the degree of shrinkage introduced is high. Moreover, we allow for leptokurtic stochastic volatility in the observation equation. The empirical findings indicate that several variants of the proposed approach outperform typical competitors frequently used in the literature, both in terms of point and density forecasts.

Estimating Stock Status and Biological Reference Points of the Sardine Fishery Using the Surplus Production Model from the Bay of Bengal, Bangladesh

This research examined the biological reference points (BRPs) and stock status of the sardine fishery in the Bay of Bengal (BoB), Bangladesh, to determine the sustainability of this resource. The Monte Carlo method (CMSY), the Bayesian state-space Schaefer surplus production model (BSM), and the ASPIC (a Stock Production Model Incorporating Covariates) software suite, were used to analyze catch–effort data obtained from the Yearbook of Fisheries Statistics of Bangladesh. All models derived maximum sustainable yields (MSY) ranging from 37,900 to 41,280 t, which is quite near to the catch from the latest year (38,051 t in 2020), indicating the fully exploited status of sardines. The estimated B &lt; BMSY and F &gt; FMSY values from the BSM and Schaefer models indicate a poor biomass and an unsafe fishing status. For Schaefer and BSM, the calculated F/FMSY values were 1.07 and 1.06, and the B/BMSY values were 0.92 and 0.75, which also indicate the overexploited status of the sardine fishery in the BoB, Bangladesh. This information will aid in developing management strategies and conservation policies for the sustainability and rebuilding of this commercially important resource in the BoB on the Bangladesh coast.

Global and local components of output gaps

This paper proposes a multi-level dynamic factor model to identify common components in output gap estimates. We pool multiple estimates for 157 countries and decompose them into one global, eight regional, and 157 country-specific cycles. Our approach easily deals with mixed frequencies, ragged edges, and discontinuities in the underlying output gap estimates. To restrict the parameter space in the Bayesian state space model, we apply a stochastic search variable selection approach and base the prior inclusion probabilities on spatial information. Our results suggest that the global and the regional cycles explain a substantial proportion of the output gaps. On average, 18% of a country’s output gap is attributable to the global cycle, 24% to the regional cycle, and 58% to the local cycle.

Identifying historical and future global change drivers that place species recovery at risk.

Ecosystem management in the face of global change requires understanding how co-occurring threats affect species and communities. Such an understanding allows for effective management strategies to be identified and implemented. An important component of this is differentiating between factors that are within (e.g. invasive predators) or outside (e.g. drought, large wildfires) of a local manager's control. In the global biodiversity hotspot of south-western Australia, small- and medium-sized mammal species are severely affected by anthropogenic threats and environmental disturbances, including invasive predators, fire, and declining rainfall. However, the relative importance of different drivers has not been quantified. We used data from a long-term monitoring program to fit Bayesian state-space models that estimated spatial and temporal changes in the relative abundance of four threatened mammal species: the woylie (Bettongia penicillata), chuditch (Dasyurus geoffroii), koomal (Trichosurus vulpecula) and quenda (Isoodon fusciventor). We then use Bayesian structural equation modelling to identify the direct and indirect drivers of population changes, and scenario analysis to forecast population responses to future environmental change. We found that habitat loss or conversion and reduced primary productivity (caused by rainfall declines) had greater effects on species' spatial and temporal population change than the range of fire and invasive predator (the red fox Vulpes vulpes) management actions observed in the study area. Scenario analysis revealed that a greater extent of severe fire and further rainfall declines predicted under climate change, operating in concert are likely to further reduce the abundance of these species, but may be mitigated partially by invasive predator control. Considering both historical and future drivers of population change is necessary to identify the factors that risk species recovery. Given that both anthropogenic pressures and environmental disturbances can undermine conservation efforts, managers must consider how the relative benefit of conservation actions will be shaped by ongoing global change.

Effects of culling vampire bats on the spatial spread and spillover of rabies virus.

Controlling pathogen circulation in wildlife reservoirs is notoriously challenging. In Latin America, vampire bats have been culled for decades in hopes of mitigating lethal rabies infections in humans and livestock. Whether culls reduce or exacerbate rabies transmission remains controversial. Using Bayesian state-space models, we show that a 2-year, spatially extensive bat cull in an area of exceptional rabies incidence in Peru failed to reduce spillover to livestock, despite reducing bat population density. Viral whole genome sequencing and phylogeographic analyses further demonstrated that culling before virus arrival slowed viral spatial spread, but reactive culling accelerated spread, suggesting that culling-induced changes in bat dispersal promoted viral invasions. Our findings question the core assumptions of density-dependent transmission and localized viral maintenance that underlie culling bats as a rabies prevention strategy and provide an epidemiological and evolutionary framework to understand the outcomes of interventions in complex wildlife disease systems.

Variational Bayesian State Space Model for dynamic process fault detection

Industrial processes are subject to various noise disturbances that lead to the stochastic nature of the modeled system and the uncertainty of the model parameters. In this paper, a variational Bayesian State Space Model (VBSSM) model is developed for dynamic process monitoring, allowing the uncertainty of parameters to be described by probabilities. The conjugate prior allows the posterior distribution of the parameters to be estimated iteratively through the Rauch Tung Striebel (RTS) smoothing estimation with the VB framework. The next state can be better predicted even in the presence of noisy perturbations, and imperceptible correlations among multiple variables can be found. By maximizing the lower bound of the objective function to approximate the true posterior distribution of the parameters, the fault detection index is constructed by the corresponding residual. Finally, the effectiveness and superiority of the proposed method is verified by numerical simulations, Tennessee Eastman process and hot rolling industry examples.