Autologistic Regression Model Research Articles

Adjusting statistical benchmark risk analysis to account for non-spatial autocorrelation, with application to natural hazard risk assessment

ABSTRACT We develop and study a quantitative, interdisciplinary strategy for conducting statistical risk analyses within the ‘benchmark risk’ paradigm of contemporary risk assessment when potential autocorrelation exists among sample units. We use the methodology to explore information on vulnerability to natural hazards across 3108 counties in the conterminous 48 US states, applying a place-based resilience index to an existing knowledgebase of hazardous incidents and related human casualties. An extension of a centered autologistic regression model is applied to relate local, county-level vulnerability to hazardous outcomes. Adjustments for autocorrelation embedded in the resiliency information are applied via a novel, non-spatial neighborhood structure. Statistical risk-benchmarking techniques are then incorporated into the modeling framework, wherein levels of high and low vulnerability to hazards are identified.

Spatial Analysis of the 2017 Outbreak of Hemorrhagic Disease and Physiographic Region in the Eastern United States.

Hemorrhagic disease (HD) is considered one of the most significant infectious diseases of white-tailed deer in North America. Investigations into environmental conditions associated with outbreaks suggest drought conditions are strongly correlated with outbreaks in some regions of the United States. However, during 2017, an HD outbreak occurred in the Eastern United States which appeared to be associated with a specific physiographic region, the Appalachian Plateau, and not drought conditions. The objective of this study was to determine if reported HD in white-tailed deer in 2017 was correlated with physiographic region. There were 456 reports of HD from 1605 counties across 26 states and 12 physiographic regions. Of the 93 HD reports confirmed by virus isolation, 76.3% (71/93) were identified as EHDV-2 and 66.2% (47/71) were from the Appalachian Plateau. A report of HD was 4.4 times more likely to occur in the Appalachian Plateau than not in 2017. Autologistic regression models suggested a statistically significant spatial dependence. The underlying factors explaining this correlation are unknown, but may be related to a variety of host, vector, or environmental factors. This unique outbreak and its implications for HD epidemiology highlight the importance for increased surveillance and reporting efforts in the future.

Assessment of sociodemographic disparities in environmental exposure might be erroneous due to neighborhood effect averaging: Implications for environmental inequality research

The neighborhood effect averaging problem (NEAP) is a major methodological problem that might affect the accuracy of assessments of individual exposure to mobility-dependent environmental factors (e.g., air/noise pollution, green/blue spaces, or healthy food environments). Focusing on outdoor ground-level ozone as a major air pollutant, this paper examines the NEAP in the evaluation of sociodemographic disparities in people’s air pollution exposures in Los Angeles using one-day activity-travel diary data of 3790 individuals. It addresses two questions: (1) How does the NEAP affect the evaluation of sociodemographic disparities in people’s air pollution exposures? (2) Which social groups with high residence-based exposures do not experience neighborhood effect averaging? The results of our spatial regression models indicate that assessments of sociodemographic disparities in people’s outdoor ground-level ozone exposures might be erroneous when people’s daily mobility is ignored because of the different manifestations of neighborhood effect averaging for different social/racial groups. The results of our spatial autologistic regression model reveal that non-workers (e.g., the unemployed, homemakers, the retired, and students) do not experience downward averaging: they have significantly lower odds of experiencing downward averaging that could have attenuated their high exposures experienced in their residential neighborhoods while traveling to other neighborhoods (thus, being doubly disadvantaged). Therefore, to avoid erroneous conclusions in environmental inequality research and ineffective public policies, it would be critical to take the NEAP into account in future studies of sociodemographic disparities related to mobility-dependent environmental factors.

A Spatial Autologistic Model to Predict the Presence of Arsenic in Private Wells Across Gaston County, North Carolina Using Geology, Well Depth, and pH

Chronic exposure to arsenic-contaminated drinking water is detrimental to human health. We develop an autologistic regression model to evaluate if the geology, pH, and well depth can improve our ability to predict the presence of arsenic at and above detectable levels (≥ 5 µg/L) found in private wells. We use arsenic samples measured in private well water across Gaston County, North Carolina, from 2011 to 2017. We use kriging to map the probability of arsenic at detectable levels across Gaston County. Arsenic at detectable levels was reported at 78 private wells. The median pH for samples containing detectable levels of arsenic was 7.3 and for samples with arsenic < 5 µg/L was 7.1. Our spatial autologistic model suggests that arsenic at detectable levels is positively associated with pH. In addition, private wells set in Mica schist (ЄZms) were associated with arsenic, suggesting a local-scale geologic source influence of arsenic in the county. Our kriging map shows that the northwestern section of the county has more than a 50% probability to have arsenic at detectable levels. In conclusion, based on our results, we recommend increased testing for wells in the Mica schist area. The map of probability of arsenic at and above detectable levels can be used to implement cost-effective targeted interventions.

Eigenvector Spatial Filtering-Based Logistic Regression for Landslide Susceptibility Assessment

Logistic regression methods have been widely used for landslide research. However, previous studies have seldom paid attention to the frequent occurrence of spatial autocorrelated residuals in regression models, which indicate a model misspecification problem and unreliable results. This study accounts for spatial autocorrelation by implementing eigenvector spatial filtering (ESF) into logistic regression for landslide susceptibility assessment. Based on a landslide inventory map and 11 landslide predisposing factors, we developed the eigenvector spatial filtering-based logistic regression (ESFLR) model, as well as a conventional logistic regression (LR) model and an autologistic regression (ALR) model for comparison. The three models were evaluated and compared in terms of their prediction capability and model fit. The ESFLR model performed better than the other two models. The overall predictive accuracy of the ESFLR model was 90.53%, followed by the ALR model (76.21%) and the LR model (74.76%), and the areas under the ROC curves for the ESFLR, ALR and LR models were 0.957, 0.828 and 0.818, respectively. The ESFLR model adequately addressed the spatial autocorrelation of residuals by reducing the Moran’s I value of the residuals to 0.0270. In conclusion, the ESFLR model is an effective and flexible method for landslide analysis.

Autologistic network model on binary data for disease progression study.

This paper focuses on analysis of spatiotemporal binary data with absorbing states. The research was motivated by a clinical study on amyotrophic lateral sclerosis (ALS), a neurological disease marked by gradual loss of muscle strength over time in multiple body regions. We propose an autologistic regression model to capture complex spatial and temporal dependencies in muscle strength among different muscles. As it is not clear how the disease spreads from one muscle to another, it may not be reasonable to define a neighborhood structure based on spatial proximity. Relaxing the requirement for prespecification of spatial neighborhoods as in existing models, our method identifies an underlying network structure empirically to describe the pattern of spreading disease. The model also allows the network autoregressive effects to vary depending on the muscles' previous status. Based on the joint distribution derived from this autologistic model, the joint transition probabilities of responses among locations can be estimated and the disease status can be predicted in the next time interval. Model parameters are estimated through maximization of penalized pseudo-likelihood. Postmodel selection inference was conducted via a bias-correction method, for which the asymptotic distributions were derived. Simulation studies were conducted to evaluate the performance of the proposed method. The method was applied to the analysis of muscle strength loss from the ALS clinical study.

Are you going to get a ticket or a warning for speeding? An autologistic regression analysis in Burlington, VT

Law enforcement is critical for improving traffic safety. However, disputes on the equity in law enforcement have continuously exacerbated the distrust between the public and the law enforcement agencies in the United States in the past decades. This study explores this issue by identifying factors influencing outcomes of traffic stops - the most common scenarios where people need to deal with law enforcement agencies. To exclude possible confounding factors, this study specifically focuses on speeding traffic stops leading to tickets or warnings in Burlington, Vermont from 2012 to 2017. The Euclidean distance-based autologistic regression model is adopted due to the presence of spatial correlations of traffic stops. It is found that with the increasing speeding severity, a speeding traffic stop is more likely to lead to a ticket. Speeding of 20 mph over the speed limit significantly influences the penalty type. Young drivers, male drivers and minority drivers are found to be more likely to be issued tickets, which suggests the possible presence of some inherent biases against these groups. In addition, time of day and month are also found to influence the likelihood of receiving speeding tickets. These findings are expected to help both the public and law enforcement agencies to better understand the characteristics of law enforcement and take appropriate measures to eliminate possible biases.

A new centered spatio-temporal autologistic regression model with an application to local spread of plant diseases

We propose a new spatio-temporal autologistic centered model for binary data on a lattice. Centering allows the self-regression coefficients to be interpreted by separating the large-scale structure from the small-scale structure. One of the coefficients determines the overall level (or average) of the process, the second determines the spatial autocorrelation. We discuss the existence of the joint distribution of the process and carry out numerical studies to highlight the interest of this type of centering. We suggest using the estimator that maximises the Pseudo Likelihood (denoted Maximum Pseudo Likelihood Estimator (MPLE) in the following) and we give a method for choosing the neighbourhood structure. We run simulations studies that show that the estimation method and model selection method work well. The method is applied to model and fit epidemiological data on Esca disease in a vineyard in the Bordeaux region.

Determinant Factors and Spatial Distribution of Visceral Leishmaniasis in North Gondar in Selected Hospitals

Introduction: Visceral Leishmaniasis (VL) is a neglected tropical disease caused by Leishmaniasis protozoa, and transmitted by a sand fly vector. Leishmaniasis is distributed in worldwide and affects millions of peoples. Objective: The objective was to assess the prevalence and spatial distribution of VL in North Gondar. Methodology: 369 patients were selected from records or medical charts retrospectively using stratified and systematic sampling techniques. The study used spatial autocorrelation measures and auto logistic regression model. Result: From 369 patients, 313 (84.8%) were infected by VL. The proportion of VL incidence for those who traveled to endemic areas was (70.2%). Male patients accounted a larger proportion (93.2%) compared to female patients (6.8%). Moran’s scatter plot test revealed there is regional clustering on the VL incidence. Local Moron’s Index value of Gondar Zuria, Dembeya, Quara, Metemma, Tach Armachiho, Gondar Town, West Belesa and Adarakay was positive, this showed places that shared boundaries have similar VL incidence (i.e high-high or low-low). From result of local Getis and Ord statistic, positive standardized Z-values of Adarkay, Gondar Town, Metema, Quara, West Belesa and Dembiya showed the places were hot spot. Similarly negative standardized Z values of Gondar Zuria and Tach Armachiho revealed that places were cold spot. The spatial auto logistic regression result revealed that sex, travel history, BMI, rainfall and elevation were the significant factors of VL at 95% confidence level. Conclusion: The prevalence of VL is still very high (84.8%) in North Gondar Zone. VL clustered on endemic districts that shared common boundaries.

Better Autologistic Regression

Autologistic regression is an important probability model for dichotomous random variables observed along with covariate information. It has been used in various fields for analyzing binary data possessing spatial or network structure. The model can be viewed as an extension of the autologistic model (also known as the Ising model, quadratic exponential binary distribution, or Boltzmann machine) to include covariates. It can also be viewed as an extension of logistic regression to handle responses that are not independent. Not all authors use exactly the same form of the autologistic regression model. Variations of the model differ in two respects. First, the variable coding---the two numbers used to represent the two possible states of the variables---might differ. Common coding choices are (zero, one) and (minus one, plus one). Second, the model might appear in either of two algebraic forms: a standard form, or a recently proposed centered form. Little attention has been paid to the effect of these differences, and the literature shows ambiguity about their importance. It is shown here that changes to either coding or centering in fact produce distinct, non-nested probability models. Theoretical results, numerical studies, and analysis of an ecological data set all show that the differences among the models can be large and practically significant. Understanding the nature of the differences and making appropriate modelling choices can lead to significantly improved autologistic regression analyses. The results strongly suggest that the standard model with plus/minus coding, which we call the symmetric autologistic model, is the most natural choice among the autologistic variants.

Autologistic models for benchmark risk or vulnerability assessment of urban terrorism outcomes.

We develop a quantitative methodology to characterize vulnerability among 132 U.S. urban centers ('cities') to terrorist events, applying a place-based vulnerability index to a database of terrorist incidents and related human casualties. A centered autologistic regression model is employed to relate urban vulnerability to terrorist outcomes and also to adjust for autocorrelation in the geospatial data. Risk-analytic 'benchmark' techniques are then incorporated into the modeling framework, wherein levels of high and low urban vulnerability to terrorism are identified. This new, translational adaptation of the risk-benchmark approach, including its ability to account for geospatial autocorrelation, is seen to operate quite flexibly in this socio-geographic setting.

Spatio-temporal clustering analysis and its determinants of hand, foot and mouth disease in Hunan, China, 2009\u20132015

BackgroundHand, foot and mouth disease (HFMD) is one of the highest reported infectious diseases with several outbreaks across the world. This study aimed at describing epidemiological characteristics, investigating spatio-temporal clustering changes, and identifying determinant factors in different clustering areas of HFMD.MethodsDescriptive statistics was used to evaluate the epidemic characteristics of HFMD from 2009 to 2015. Spatial autocorrelation and spatio-temporal cluster analysis were used to explore the spatial temporal patterns. An autologistic regression model was employed to explore determinants of HFMD clustering.ResultsThe incidence rates of HFMD ranged from 54.31/10 million to 318.06/10 million between 2009 and 2015 in Hunan. Cases were mainly prevalent in children aged 5 years and even younger, with an average male-to-female sex ratio of 1.66, and two epidemic periods in each year. Clustering areas gathered in the northern regions in 2009 and in the central regions from 2010 to 2012. They moved to central-southern regions in 2013 and 2014 and central-western regions in 2015. The significant risk factors of HFMD clusters were rainfall (OR = 2.187), temperature (OR = 4.329) and humidity (OR = 2.070). The protect factor was wind speed (OR = 0.258).ConclusionsThe HFMD incidence from 2009 to 2015 in Hunan showed a new spatiotemporal clustering tendency, with the shifting trend of clustering areas toward south and west. Meteorological factors showed a strong association with HFMD clustering, which may assist in predicting future spatial-temporal clusters.

Spatial Patterns and Impacts of Environmental and Climatic Factors on Canine Sinonasal Aspergillosis in Northern California.

Sinonasal aspergillosis (SNA) causes chronic nasal discharge in dogs and has a worldwide distribution, although most reports of SNA in North America originate from the western USA. SNA is mainly caused by Aspergillus fumigatus, a ubiquitous saprophytic filamentous fungus. Infection is thought to follow inhalation of spores. SNA is a disease of the nasal cavity and/or sinuses with variable degrees of local invasion and destruction. While some host factors appear to predispose to SNA (such as belonging to a dolichocephalic breed), environmental risk factors have been scarcely studied. Because A. fumigatus is also the main cause of invasive aspergillosis in humans, unraveling the distribution and the environmental and climatic risk factors for this agent in dogs would be of great benefit for public health studies, advancing understanding of both distribution and risk factors in humans. In this study, we reviewed electronic medical records of 250 dogs diagnosed with SNA between 1990 and 2014 at the University of California Davis Veterinary Medical Teaching Hospital (VMTH). A 145-mile radius catchment area around the VMTH was selected. Data were aggregated by zip code and incorporated into a multivariate logistic regression model. The logistic regression model was compared to an autologistic regression model to evaluate the effect of spatial autocorrelation. Traffic density, active composting sites, and environmental and climatic factors related with wind and temperature were significantly associated with increase in disease occurrence in dogs. Results provide valuable information about the risk factors and spatial distribution of SNA in dogs in Northern California. Our ultimate goal is to utilize the results to investigate risk-based interventions, promote awareness, and serve as a model for further studies of aspergillosis in humans.

Erratum to: On Estimation and Selection of Autologistic Regression Models via Penalized Pseudolikelihood

Erratum to: On Estimation and Selection of Autologistic Regression Models via Penalized Pseudolikelihood

Classification of Large-Scale Remote Sensing Images for Automatic Identification of Health Hazards

Remote sensing images from Earth-orbiting satellites are a potentially rich data source for monitoring and cataloguing atmospheric health hazards that cover large geographic regions. A method is proposed for classifying such images into hazard and nonhazard regions using the autologistic regression model, which may be viewed as a spatial extension of logistic regression. The method includes a novel and simple approach to parameter estimation that makes it well suited to handling the large and high-dimensional datasets arising from satellite-borne instruments. The methodology is demonstrated on both simulated images and a real application to the identification of forest fire smoke.

Simulation of urban agglomeration ecosystem spatial distributions under different scenarios: A case study of the Changsha–Zhuzhou–Xiangtan urban agglomeration

Changsha–Zhuzhou–Xiangtan is an urban agglomeration along the middle reaches of the Yangtze River, the development plan of which is listed as one of national strategic development. The eco-environmental quality of the Changsha–Zhuzhou–Xiangtan urban agglomeration concerns the ecological security of China. It is essential to discuss the method of creating urban agglomeration plans to increase ecological security. In this paper, the ecosystem distributions were mapped using the CLUE-S model incorporated with Gray Model (GM) (1, 1) and an auto-logistic regression model under the conditions of a natural increase scenario (NIS), a cultivated protection scenario (CPS) and an ecological protection scenario (EPS). We analyzed the change and conversion characteristics of ecosystems both in the study area and the junction of Changsha, Zhuzhou and Xiangtan. The results showed that the ecosystem change model developed in this paper performed well in mapping future urban ecosystem distributions. The change of the ecosystem spatial distributions showed us that the built-up ecosystem would expand in the future by transforming cultivated and green land ecosystems and that the boundaries between cities would be blurred. In the whole study area, the amount of converted area from cultivated and green land to built-up land was lower in CPS and EPS than in NIS. However, in the key area, the converted result was contrary to the whole study area. Although the CPS and EPS were beneficial to the eco-environmental protection of the whole study region, the sub-regional eco-environment should be given more attention to assure that the ecological security of the whole region remains safe.

A random effect autologistic regression model with application to the characterization of multiple microstructure samples

ABSTRACTThe microstructure of a material can strongly influence its properties such as strength, hardness, wear resistance, etc., which in turn play an important role in the quality of products produced from these materials. Existing studies on a material's microstructure have mainly focused on the characteristics of a single microstructure sample and the variation between different microstructure samples is ignored. In this article, we propose a novel random effect autologistic regression model that can be used to characterize the variation in microstructures between different samples for two-phase materials that consist of two distinct parts with different chemical structures. The proposed model differs from the classic autologistic regression model in that we consider the unit-to-unit variability among the microstructure samples, which is characterized by the random effect parameters. To estimate the model parameters given a set of microstructure samples, we first derive a likelihood function, based on which a maximum likelihood estimation method is developed. However, maximizing the likelihood function of the proposed model is generally difficult as it has a complex form. To overcome this challenge, we further develop a stochastic approximation expectation maximization algorithm to estimate the model parameters. A simulation study is conducted to verify the proposed methodology. A real-world example of a dual-phase high strength steel is used to illustrate the developed methods.

Simulating urban land use change by incorporating an autologistic regression model into a CLUE-S model

The Conversion of Land Use and its Effects at Small regional extent (CLUE-S) model is a widely used method to simulate land use change. An ordinary logistic regression model was integrated into the CLUE-S model to identify explanatory variables without considering the spatial autocorrelation effect. Using image-derived maps of the Changsha-Zhuzhou-Xiangtan urban agglomeration, the CLUE-S model was integrated with the ordinary logistic regression and autologistic regression models in this paper to simulate land use change in 2000, 2005 and 2009 based on an observation map from 1995. Significant positive spatial autocorrelation was detected in residuals of ordinary logistic models. Some variables that were much more significant than they should be were selected. Autologistic regression models, which used autocovariate incorporation, were better able to identify driving factors. The Receiver Operating Characteristic Curve (ROC) values of autologistic regression models were larger than 0.8 and the pseudo R2 values were improved, compared with results of logistic regression model. By overlapping the observation maps, the Kappa values of the ordinary logistic regression model (OL)-CLUE-S and autologistic regression model (AL)-CLUE-S models were larger than 0.75. The results showed that the simulation results were indeed accurate. The Kappa fuzzy (Kfuzzy) values of the AL-CLUE-S models (0.780, 0.773, 0.606) were larger than the values of the OL-CLUE-S models (0.759, 0.760, 0.599) during the three periods. The AL-CLUE-S models performed better than the OL-CLUE-S models in the simulation of land use change. The results showed that it is reasonable to integrate autocovariates into CLUE-S models. However, the Kfuzzy values decreased with prolonged duration of simulation and the maximum range of time was not discussed in this paper.

Tsetse fly (G. f. fuscipes) distribution in the Lake Victoria basin of Uganda.

Tsetse flies transmit trypanosomes, the causative agent of human and animal African trypanosomiasis. The tsetse vector is extensively distributed across sub-Saharan Africa. Trypanosomiasis maintenance is determined by the interrelationship of three elements: vertebrate host, parasite and the vector responsible for transmission. Mapping the distribution and abundance of tsetse flies assists in predicting trypanosomiasis distributions and developing rational strategies for disease and vector control. Given scarce resources to carry out regular full scale field tsetse surveys to up-date existing tsetse maps, there is a need to devise inexpensive means for regularly obtaining dependable area-wide tsetse data to guide control activities. In this study we used spatial epidemiological modelling techniques (logistic regression) involving 5000 field-based tsetse-data (G. f. fuscipes) points over an area of 40,000 km2, with satellite-derived environmental surrogates composed of precipitation, temperature, land cover, normalised difference vegetation index (NDVI) and elevation at the sub-national level. We used these extensive tsetse data to analyse the relationships between presence of tsetse (G. f. fuscipes) and environmental variables. The strength of the results was enhanced through the application of a spatial autologistic regression model (SARM). Using the SARM we showed that the probability of tsetse presence increased with proportion of forest cover and riverine vegetation. The key outputs are a predictive tsetse distribution map for the Lake Victoria basin of Uganda and an improved understanding of the association between tsetse presence and environmental variables. The predicted spatial distribution of tsetse in the Lake Victoria basin of Uganda will provide significant new information to assist with the spatial targeting of tsetse and trypanosomiasis control.

Bayesian Inference of Spatially Correlated Binary Data Using Skew-Normal Latent Variables with Application in Tooth Caries Analysis

The analysis of spatially correlated binary data observed on lattices is an interesting topic that catches the attention of many scholars of different scientific fields like epidemiology, medicine, agriculture, biology, geology and geography. To overcome the encountered difficulties upon fitting the autologistic regression model to analyze such data via Bayesian and/or Markov chain Monte Carlo (MCMC) techniques, the Gaussian latent variable model has been enrolled in the methodology. Assuming a normal distribution for the latent random variable may not be realistic and wrong, normal assumptions might cause bias in parameter estimates and affect the accuracy of results and inferences. Thus, it entails more flexible prior distributions for the latent variable in the spatial models. A review of the recent literature in spatial statistics shows that there is an increasing tendency in presenting models that are involving skew distributions, especially skew-normal ones. In this study, a skew-normal latent variable modeling was developed in Bayesian analysis of the spatially correlated binary data that were acquired on uncorrelated lattices. The proposed methodology was applied in inspecting spatial dependency and related factors of tooth caries occurrences in a sample of students of Yasuj University of Medical Sciences, Yasuj, Iran. The results indicated that the skew-normal latent variable model had validity and it made a decent criterion that fitted caries data.