Search

Obtaining a representative sample of disease vectors (mosquitoes, flies, ticks, etc.) is essential for researchers to draw meaningful conclusions about the entire vector population in a target study area and during a specific study period. To achieve this, a carefully chosen surveillance design is required to ensure that the sample captures essential spatial and temporal variations in the target vector population(s) and/or that the study results can be generalized to the entire population. Designed-based and model-based spatiotemporal sampling (or in our context surveillance) designs can be used to maximize information gain within given resource constraints. In this paper, we aim to offer a concise overview of common spatiotemporal field sampling designs, their advantages and disadvantages and their practical applications in the context of surveillance and management of vector-borne diseases. At the end of the article, we offer guidance to help vector-borne disease surveillance planners design effective spatiotemporal surveillance interventions.

Abstract This paper provides new tools for analyzing spatio-temporal event networks. We build time series of directed event networks for a set of spatial distances, and based on scan-statistics, the spatial distance that generates the strongest change of event network connections is chosen. In addition, we propose an empirical random network event generator to detect significant motifs throughout time. This generator preserves the spatial configuration but randomizes the order of the occurrence of events. To prevent the large number of links from masking the count of motifs, we propose using standardized counts of motifs at each time slot. Our methodology is able to detect interaction radius in space, build time series of networks, and describe changes in its topology over time, by means of identification of different types of motifs that allows for the understanding of the spatio-temporal dynamics of the phenomena. We illustrate our methodology by analyzing thefts occurred in Medellín (Colombia) between the years 2003 and 2015.

Crime activity in many cities worldwide causes significant damages to the lives of victims and their surrounding communities. It is a public disorder problem, and big cities experience large amounts of crime events. Spatio-temporal prediction of crimes activity can help the cities to have a better allocation of police resources and surveillance. Deep learning techniques are considered efficient tools to predict future events analyzing the behavior of past ones; however, they are not usually applied to crime event prediction using a spatio-temporal approach. In this paper, a Convolutional Neural Network (CNN) together with a Long-Short Term Memory (LSTM) network (thus CLSTM-NN) are proposed to predict the presence of crime events over the city of Baltimore (USA). In particular, matrices of past crime events are used as input to a CLSTM-NN to predict the presence of at least one event in future days. The model is implemented on two types of events: “street robbery” and “larceny”. The proposed procedure is able to take into account spatial and temporal correlations present in the past data to improve future prediction. The prediction performance of the proposed neural network is assessed under a number of controlled plausible scenarios, using some standard metrics (Accuracy, AUC-ROC, and AUC-PR).

Emergency calls are defined by an ever-expanding utilisation of information and sensing technology, leading to extensive volumes of spatio-temporal high-resolution data. The spatial and temporal character of the emergency calls is leveraged by authorities to allocate resources and infrastructure for an effective response, to identify high-risk event areas, and to develop contingency strategies. In this context, the spatio-temporal analysis of emergency calls is crucial to understanding and mitigating distress situations. However, modelling and predicting crime-related emergency calls remain challenging due to their heterogeneous and dynamic nature with complex underlying processes. In this context, we propose a modelling strategy that accounts for the intrinsic complex space–time dynamics of some crime data on cities by handling complex advection, diffusion, relocation, and volatility processes. This study presents a predictive framework capable of assimilating data and providing confidence estimates on the predictions. By analysing the dynamics of the weekly number of emergency calls in Valencia, Spain, for ten years (2010–2020), we aim to understand and forecast the spatio-temporal behaviour of emergency calls in an urban environment. We include putative geographical variables, as well as distances to relevant city landmarks, into the spatio-temporal point process modelling framework to measure the effect deterministic components exert on the intensity of emergency calls in Valencia. Our results show how landmarks attract or repel offenders and act as proxies to identify areas with high or low emergency calls. We are also able to estimate the weekly average growth and decay in space and time of the emergency calls. Our proposal is intended to guide mitigation strategies and policy.

Crime is a negative phenomenon that affects the daily life of the population and its development. When modeling crime data, assumptions on either the spatial or the temporal relationship between observations are necessary if any statistical analysis is to be performed. In this paper, we structure space–time dependency for count data by considering a stochastic difference equation for the intensity of the space–time process rather than placing structure on a latent space–time process, as Cox processes would do. We introduce a class of spatially correlated self-exciting spatio-temporal models for count data that capture both dependence due to self-excitation, as well as dependence in an underlying spatial process. We follow the principles in Clark and Dixon (2021) but considering a generalized additive structure on spatio-temporal varying covariates. A Bayesian framework is proposed for inference of model parameters. We analyze three distinct crime datasets in the city of Riobamba (Ecuador). Our model fits the data well and provides better predictions than other alternatives.

The brown rat has been living with humans in a wide variety of environmental contexts; it adversely affects public health by transmission of pathogens that can cause human diseases and allergies. Understanding behavioral aspects and environmental factors of pest species can contribute to their effective management and control. This aim of this study was to investigate the spatial distribution of rats in Latina district of Madrid (Spain), and its relationship with several potential multi-source foci. A focus is any particular social and environmental urban scenario that favors the clustering and proliferation of rats. We have developed a statistical framework to provide valid information on the spatial distribution and behavior of the rats around identified potential foci that favor the concentration of rats in urban environments. We extended the standard Poisson regression model by the inclusion of a multiplicative non-linear function of the distance, an unstructured random effect, and a spatial random effect to account for the spatial structure of socio-demographic and environmental covariates were also considered in the model to control for potential confusion. We found evidence of an association between the spatial distribution of rats aggregated by census tracts and distance to foci, and this association was controlled by the covariates considered in this study. Keywords: Environmental sources, Focus, Rat sightings, Rattus norvegicus, Spatial Poisson regression

The doubly stochastic mechanism generating the realizations of spatial log-Gaussian Cox processes is empirically assessed in terms of generalized entropy, divergence and complexity measures. The aim is to characterize the contribution to stochasticity from the two phases involved, in relation to the transfer of information from the intensity field to the resulting point pattern, as well as regarding their marginal random structure. A number of scenarios are explored regarding the Matérn model for the covariance of the underlying log-intensity random field. Sensitivity with respect to varying values of the model parameters, as well as of the deformation parameters involved in the generalized informational measures, is analyzed on the basis of regular lattice partitionings. Both a marginal global assessment based on entropy and complexity measures, and a joint local assessment based on divergence and relative complexity measures, are addressed. A Poisson process and a log-Gaussian Cox process with white noise intensity, the first providing an upper bound for entropy, are considered as reference cases. Differences regarding the transfer of structural information from the intensity field to the subsequently generated point patterns, reflected by entropy, divergence and complexity estimates, are discussed according to the specifications considered. In particular, the magnitude of the decrease in marginal entropy estimates between the intensity random fields and the corresponding point patterns quantitatively discriminates the global effect of the additional source of variability involved in the second phase of the double stochasticity.

The aim of this work was to extend the results of Perez et al. (Physica A (2006), 365 (2), 282−288) to the two-dimensional (2D) fractional Brownian field. In particular, we defined Shannon entropy using the wavelet spectrum from which the Hurst exponent is estimated by the regression of the logarithm of the square coefficients over the levels of resolutions. Using the same methodology. we also defined two other entropies in 2D: Tsallis and the Rényi entropies. A simulation study was performed for showing the ability of the method to characterize 2D (in this case, α = 2 ) self-similar processes.

Cancer is a widely spread disease that affects a large proportion of the human population, and many research teams are developing algorithms to help medics to understand this disease. In particular, tumor growth has been studied from different viewpoints and several mathematical models have been proposed. In this paper, we review a set of comprehensive and modern tools that are useful for prediction of cancer growth in space and time. We comment on three alternative approaches. We first consider spatio-temporal stochastic processes within a Bayesian framework to model spatial heterogeneity, temporal dependence and spatio-temporal interactions amongst the pixels, providing a general modeling framework for such dynamics. We then consider predictions based on geometric properties of plane curves and vectors, and propose two methods of geometric prediction. Finally we focus on functional data analysis to statistically compare tumor contour evolutions. We also analyze real data on brain tumor.

We propose a wavelet-based approach to measure the Shannon entropy in the context of spatial point patterns. The method uses the fully anisotropic Morlet wavelet to estimate the energy distribution at different directions and scales. The spatial heterogeneity and complexity of spatial point patterns is then analyzed using the multiscale anisotropic wavelet entropy. The efficacy of the approach is shown through a simulation study. Finally, an application to the catalog of earthquake events in Chile is considered.

The explosion of data in the information age has provided an opportunity to explore the possibility of characterizing the climate patterns using data mining techniques. Nigeria has a unique tropical climate with two precipitation regimes: low precipitation in the north leading to aridity and desertification and high precipitation in parts of the southwest and southeast leading to large scale flooding. In this research, four indices have been used to characterize the intensity, frequency, and amount of rainfall over Nigeria. A type of Artificial Neural Network called the self-organizing map has been used to reduce the multiplicity of dimensions and produce four unique zones characterizing extreme precipitation conditions in Nigeria. This approach allowed for the assessment of spatial and temporal patterns in extreme precipitation in the last three decades. Precipitation properties in each cluster are discussed. The cluster closest to the Atlantic has high values of precipitation intensity, frequency, and duration, whereas the cluster closest to the Sahara Desert has low values. A significant increasing trend has been observed in the frequency of rainy days at the center of the northern region of Nigeria.

This paper represents a survey of recent advances in modeling of space or space-time Gaussian Random Fields (GRF), tools of Geostatistics at hand for the understanding of special cases of noise in image analysis. They can be used when stationarity or isotropy are unrealistic assumptions, or even when negative covariance between some couples of locations are evident. We show some strategies in order to escape from these restrictions, on the basis of rich classes of well known stationary or isotropic non negative covariance models, and through suitable operations, like linear combinations, generalized means, or with particular Fourier transforms.

We describe a simple way to construct new statistical models for spatial point pattern data. Taking two or more existing models (finite Gibbs spatial point processes) we multiply the probability densities together and renormalise to obtain a new probability density. We call the resulting model a hybrid. We discuss stochastic properties of hybrids, their statistical implications, statistical inference, computational strategies and software implementation in the R package spatstat. Hybrids are particularly useful for constructing models which exhibit interaction at different spatial scales. The methods are demonstrated on a real data set on human social interaction. Software and data are provided.

Spatially correlated curves are present in a wide range of applied disciplines. In this paper we describe the R package geofd which implements ordinary kriging prediction for this type of data. Initially the curves are pre-processed by fitting a Fourier or B-splines basis functions. After that the spatial dependence among curves is estimated by means of the trace-variogram function. Finally the parameters for performing prediction by ordinary kriging at unsampled locations are by estimated solving a linear system based estimated trace-variogram. We illustrate the software analyzing real and simulated data.Curvas espacialmente correlacionadas están presentes en un amplio rango de disciplinas aplicadas. En este trabajo se describe el paquete R geofd que implementa predicción por kriging ordinario para este tipo de datos. Inicialmente las curvas son suavizadas usando bases de funciones de Fourier o B-splines. Posteriormente la dependencia espacial entre las curvas es estimada por la función traza-variograma. Finalmente los parámetros del predictor kriging ordinario son estimados resolviendo un sistema de ecuaciones basado en la estimación de la función traza-variograma. Se ilustra el paquete analizando datos reales y simulados.