Article
Agriculture, Multidisciplinary
Michael J. Wellington, Roger Lawes, Petra Kuhnert
Summary: Satellite imagery allows for the inference of crop production trends over space and time, but the large size of these datasets poses computational challenges for statistical modeling. Recent advances in computational techniques have enabled the use of generalized additive models on very large datasets. We propose a framework using generalized additive models to infer trends in crop production, considering inter-annual trends, spatial distribution, crop ontogeny, inter-annual changes in seasonality, and inter-annual changes in spatial distribution. Application of this model to agricultural sites in Australia and Madagascar yielded valuable insights into crop production and food security.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2023)
Article
Computer Science, Interdisciplinary Applications
Giampiero Marra, Alessio Farcomeni, Rosalba Radice
Summary: The method presented can handle various types of censoring data and flexibly estimate multiple covariate effects. Through simulation and original data example, a peak of risk in the first week of hospitalization and a non-linear effect of Model for End-Stage Liver Disease (MELD) score were identified.
COMPUTATIONAL STATISTICS & DATA ANALYSIS
(2021)
Article
Mathematical & Computational Biology
Paolo Girardi, Luca Greco, Laura Ventura
Summary: This study proposes a method to fit multiple waves in a model, discussing an approach based on a change-point model in a pseudo-likelihood framework that considers model misspecification issues, and applies it to data collected in Italy.
BIOMETRICAL JOURNAL
(2022)
Article
Environmental Studies
Audrone Virbickaite, Hoang Nguyen, Minh-Ngoc Tran
Summary: This study explores the benefits of incorporating fat-tailed innovations, asymmetric volatility response, and an extended information set into crude oil return modeling and forecasting. The findings indicate that the inclusion of exogenous variables is beneficial for GARCH-type models while offering only a marginal improvement for GAS and SV-type models. Notably, GAS-family models exhibit superior performance in terms of in-sample fit, out-of-sample forecast accuracy, as well as Value-at-Risk and Expected Shortfall prediction.
Article
Statistics & Probability
Nan-Jung Hsu, Hsin-Cheng Huang, Ruey S. Tsay
Summary: Matrix-variate time series, commonly seen in various sciences, are characterized by large matrix dimensions. A structured autoregressive model is introduced to capture temporal dynamics in a matrix-variate time series, reducing dimensionality and highlighting dynamic interactions among columns and rows in the AR matrices. Incorporating spatial information and exploring sparsity in AR coefficients enhance the model's flexibility and parsimonious parameterization.
JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS
(2021)
Article
Computer Science, Theory & Methods
Xinmin Li, Haozhe Liang, Wolfgang Haerdle, Hua Liang
Summary: We propose test statistics based on the penalized spline to decide between generalized linear models and generalized partially linear models. The numerical performance of the proposed statistics is comparable to that of their kernel-based competitors, which have been shown to be asymptotically normal in the literature (Hardle et al. in J Am Stat Assoc 93:1461-1474, 1998). We also numerically explore the possibility of using the proposed statistics for goodness of fit checking for GLM. The proposed procedures are illustrated to analyze two datasets.
STATISTICS AND COMPUTING
(2023)
Article
Mathematics
Gurami Tsitsiashvili, Marina Osipova
Summary: This study estimated the ability of a particle flow model to form a copy of an image and the distance between the copy and the image using a special probability metric. The research also investigated the flowability of grinding balls on a surface using formulas of stochastic geometry. Additionally, the analysis of reconstructing characteristics of an inhomogeneous Poisson flow with inaccurate observations was conducted using the Poisson flow point colouring theorem. The dependence of the Poisson parameter on the peak load created by an inhomogeneous input flow in a queuing system was also estimated.
Article
Ecology
Nicholas J. Clark, Konstans Wells
Summary: Generalised additive models (GAMs) are popular tools for estimating smooth nonlinear relationships between predictors and response variables, but less useful for producing forecasts. Dynamic generalised additive models (DGAMs) address this limitation by jointly estimating the GAM linear predictor and unobserved dynamic components for time series. These models are particularly useful for analysing multiple series and learning complex temporal associations. mvgam R package implements these models and offers additional features for calculating correlations, model selection, online data augmentation, and visualising uncertainties. DGAMs offer a solution for forecasting discrete time series while estimating nonlinear predictor associations, which is important in applied ecology.
METHODS IN ECOLOGY AND EVOLUTION
(2023)
Article
Geosciences, Multidisciplinary
Alessandro Celani, Paolo Giudici
Summary: We propose an endemic-epidemic model, a negative binomial space-time autoregression, for monitoring the contagion dynamics of COVID-19 pandemic in both time and space. An empirical analysis of heavily affected provinces in northern Italy with similar non-pharmaceutical policy interventions is conducted to illustrate the model's applicability.
SPATIAL STATISTICS
(2022)
Article
Statistics & Probability
Nicoletta D'Angelo, Giada Adelfio, Antonino Abbruzzo, Jorge Mateu
Summary: This study analyzes the spatio-temporal distribution of visitors' stops at tourist attractions in Palermo using stochastic point process theory on linear networks. The researchers propose an inhomogeneous Poisson point process model and a Gibbs point process model with mixed effects to account for spatial interaction and clustering. A more complex spatio-temporal log-Gaussian Cox process model is also formulated to address the problem of distance metric choice.
ANNALS OF APPLIED STATISTICS
(2022)
Article
Green & Sustainable Science & Technology
Alessandra Gaeta, Gianluca Leone, Alessandro Di Menno di Bucchianico, Mariacarmela Cusano, Raffaela Gaddi, Armando Pelliccioni, Maria Antonietta Reatini, Annalisa Di Bernardino, Giorgio Cattani
Summary: High-resolution measurements of ultrafine particle concentrations in ambient air are crucial for studying the health effects of long-term exposure. This study conducted in Rome aimed to extend knowledge on small-scale spatio-temporal variability of Particle Number Concentration (PNC) through measurements in the university district, showing winter PNC values were about twice summer values and PNC in the university area was significantly lower than in external routes, with GAMs models capturing spatial variability effectively.
Correction
Statistics & Probability
Reza Azimi, Mahdy Esmailian
Summary: This note points out the mistakes made by Gomes et al. in presenting a new family of distributions and corrects them.
COMMUNICATIONS IN STATISTICS-THEORY AND METHODS
(2023)
Article
Engineering, Environmental
Andre Victor Ribeiro Amaral, Jonatan A. Gonzalez, Paula Moraga
Summary: Infectious disease modeling is crucial for understanding disease spreading and controlling it. This paper proposes a spatio-temporal modeling framework that integrates the SIR compartment and log-Gaussian Cox process (LGCP) models to characterize infectious disease dynamics. The method is validated using simulation with real and synthetic data in Sao Paulo, Brazil and applied to analyze COVID-19 dynamics in Cali, Colombia. The results demonstrate that the modified LGCP model, which incorporates information from the SIR model, outperforms equivalent models that do not consider this information. Furthermore, the proposed method allows for the incorporation of age-stratified contact information, providing valuable insights for decision-making.
STOCHASTIC ENVIRONMENTAL RESEARCH AND RISK ASSESSMENT
(2023)
Article
Green & Sustainable Science & Technology
Yi Fang, Hao Li, Yu Li, Guoxing Chen, Yuejun Lv, Yanju Peng
Summary: V-s30, widely used in seismic engineering, is rarely tested to a depth of 30 m due to limitations. Existing models for predicting Vs30 in Tangshan have deviations, and three new models are proposed in this study. The linear model is found to be more suitable for depths <= 18 m, while the conditional independent model performs well for depths > 18 m.
Correction
Statistics & Probability
Reza Azimi, Mahdy Esmailian
Summary: Pararai et al. introduced a new four-parameter distribution and made some mistakes in its presentation, which are corrected in this note.
COMMUNICATIONS IN STATISTICS-THEORY AND METHODS
(2023)
Article
Dentistry, Oral Surgery & Medicine
Nancy Lyn Wilson Westmark, Herve Sroussi, Ibon Tamayo, Alessandro Villa
Article
Statistics & Probability
Abdollah Jalilian, Jorge Mateu
Summary: This paper introduces a method using deep convolutional neural networks and a Siamese framework to distinguish structural differences between spatial point patterns. The adequacy and generality of this method is demonstrated through simulation study and data analysis.
ADVANCES IN DATA ANALYSIS AND CLASSIFICATION
(2023)
Article
Oncology
David Reparaz, Marta Ruiz, Diana Llopiz, Leyre Silva, Enric Vercher, Belen Aparicio, Josune Egea, Ibon Tamayo-Uria, Sandra Hervas-Stubbs, Jorge Garcia-Balduz, Carla Castro, Mercedes Inarrairaegui, Maria Tagliamonte, Angela Mauriello, Beatrice Cavalluzzo, Luigi Buonaguro, Charlotte Rohrer, Kathrin Heim, Catrin Tauber, Maike Hofmann, Robert Thimme, Bruno Sangro, Pablo Sarobe
Summary: Immunogenic neoantigens with potential applicability for future combinatorial therapeutic strategies may be generated in hepatocellular carcinoma tumors.
JOURNAL FOR IMMUNOTHERAPY OF CANCER
(2022)
Article
Ecology
Carles Mila, Jorge Mateu, Edzer Pebesma, Hanna Meyer
Summary: This study proposes a new cross-validation strategy that takes into account the geographical prediction space and compares it with other established methods. The new method, called NNDM LOO CV, provides reliable estimates in all scenarios considered. The existing methods, LOO and bLOO CV, have limitations and only provide accurate estimates in certain situations. Therefore, considering the geographical prediction space is essential when designing map validation methods.
METHODS IN ECOLOGY AND EVOLUTION
(2022)
Article
Genetics & Heredity
Suzanne Vogelezang, Jonathan P. Bradfield, Struan F. A. Grant, Janine F. Felix, Vincent W. V. Jaddoe
Summary: This study found associations between head circumference and intelligence, and highlighted the significant overlap of biological processes between early-life and adult head circumference. It also revealed the genetic correlations of early-life head circumference with intracranial volume, years of schooling, childhood and adult intelligence.
BMC MEDICAL GENOMICS
(2022)
Article
Statistics & Probability
Somnath Chaudhuri, Pablo Juan, Jorge Mateu
Summary: Using accident records in an urban environment, this study develops a spatio-temporal model to predict the number of traffic collisions and generate risk maps for the entire road network. The use of SPDE network triangulation to estimate spatial autocorrelation on a linear network is a novel approach. The resulting risk maps offer valuable information for accident prevention and interdisciplinary road safety measures.
JOURNAL OF APPLIED STATISTICS
(2023)
Article
Medicine, General & Internal
Jaime Gallego Perez-Larraya, Marc Garcia-Moure, Sara Labiano, Ana Patino-Garcia, Jessica Dobbs, Marisol Gonzalez-Huarriz, Marta Zalacain, Lucia Marrodan, Naiara Martinez-Velez, Montserrat Puigdelloses, Virginia Laspidea, Itziar Astigarraga, Blanca Lopez-Ibor, Ofelia Cruz, Miren Oscoz Lizarbe, Sandra Hervas-Stubbs, Gorka Alkorta-Aranburu, Ibon Tamayo, Beatriz Tavira, Ruben Hernandez-Alcoceba, Chris Jones, Gitanjali Dharmadhikari, Cristian Ruiz-Moreno, Henk Stunnenberg, Esther Hulleman, Jasper van der Lugt, Miguel A. Idoate, Ricardo Diez-Valle, Ines Esparragosa Vazquez, Maria Villalba, Carlos de Andrea, Jorge M. Nunez-Cordoba, Brett Ewald, Joan Robbins, Juan Fueyo, Candelaria Gomez-Manzano, Frederick F. Lang, Sonia Tejada, Marta M. Alonso
Summary: This study investigated the use of oncolytic virus DNX-2401 in pediatric patients with diffuse intrinsic pontine glioma (DIPG). The results showed that treatment with DNX-2401 led to changes in T-cell activity and reduction or stabilization of tumor size in some patients, but also caused adverse events.
NEW ENGLAND JOURNAL OF MEDICINE
(2022)
Article
Engineering, Environmental
Julia Calatayud, Marc Jornet, Jorge Mateu
Summary: In this study, we developed and calibrated stochastic continuous models to capture crime dynamics in the city of Valencia, Spain. By decomposing the monthly time series into trend and seasonal components, we modeled the former using geometric Brownian motions and the latter using randomly perturbed sine-cosine waves. The models, although simple, demonstrated high ability to simulate real data and showed promising potential for identifying crimes-interaction and short-term predictive policing.
STOCHASTIC ENVIRONMENTAL RESEARCH AND RISK ASSESSMENT
(2023)
Article
Mathematics, Applied
Julia Calatayud, Marc Jornet, Jorge Mateu
Summary: We study the dynamics of abstract models for crime evolution, taking into account participation in crime and incarceration. Individuals transition between three segments, and crime is viewed as a social epidemic. The models incorporate spatial variability using discrete and continuous forms of space, and the effect of the basic reproduction number on the long-term dynamics of crime is examined.
MATHEMATICAL METHODS IN THE APPLIED SCIENCES
(2023)
Article
Mathematical & Computational Biology
David Payares-Garcia, Jorge Mateu, Wiebke Schick
Summary: MRI plays an important role in the diagnosis and prognosis of neurodegenerative diseases, and can be used for accurate and automated classification. This article proposes a classification technique that incorporates uncertainty and spatial information to distinguish between healthy individuals and patients with neurodegenerative diseases. The experimental results demonstrate that including a spatially informed MRI scan increases the accuracy of classification by 25%.
STATISTICS IN MEDICINE
(2022)
Article
Engineering, Environmental
Julia Calatayud, Marc Jornet, Jorge Mateu
Summary: We propose a methodology for quantitatively fitting and forecasting real spatio-temporal crime data using stochastic differential equations. The study focuses on Valencia, Spain, using 90247 robbery and theft incidents recorded from the 112-emergency phone over eleven years (2010-2020). The incidents are categorized into 26 zip codes, and monthly crime time series are created for each zip code. By modeling the annual trend components using Ito diffusion with correlated noises, this study can simulate spatio-temporal situations and identify risky areas and periods based on present and past data.
STOCHASTIC ENVIRONMENTAL RESEARCH AND RISK ASSESSMENT
(2023)
Article
Engineering, Environmental
Asael Fabian Martinez, Somnath Chaudhuri, Carlos Diaz-Avalos, Pablo Juan, Jorge Mateu, Ramses H. Mena
Summary: An unsupervised classification method is proposed for point events occurring on a geometric network. It utilizes the flexibility and practicality of random partition models to discover clustering structures of observations from a specific phenomenon on a given set of edges. By incorporating spatial effects through a random partition distribution induced by a Dirichlet process, the method offers an appealing clustering approach. A Gibbs sampler algorithm is proposed and evaluated with sensitivity analysis. The analysis of crime and violence patterns in Mexico City serves as the motivation and illustration for this proposal.
STOCHASTIC ENVIRONMENTAL RESEARCH AND RISK ASSESSMENT
(2023)
Article
Computer Science, Interdisciplinary Applications
I. Fuentes-Santos, W. Gonzalez-Manteiga, J. Mateu
Summary: This work focuses on kernel estimation of the pair correlation function (PCF) for inhomogeneous spatial point processes. We propose a bootstrap bandwidth selector based on minimizing the mean integrated squared error (MISE). The variance term is estimated by nonparametric bootstrap, and the bias by a plug-in approach using a pilot estimator of the PCF. Kernel estimators of the PCF also require a pilot estimator of the first-order intensity. We test the performance of the bandwidth selector and the role of the pilot intensity estimator in a simulation study. The bootstrap bandwidth selector is competitive with cross-validation procedures, but the contribution of the bandwidth parameter to the goodness-of-fit of the kernel PCF estimator is minor in comparison with that of the pilot intensity function. The data-based kernel intensity estimator leads to biased kernel PCF estimators, while both kernel and parametric covariate-based intensities provide accurate estimators of the PCF.
JOURNAL OF STATISTICAL COMPUTATION AND SIMULATION
(2023)
Article
Social Sciences, Mathematical Methods
Andrea Gilardi, Riccardo Borgoni, Luca Presicce, Jorge Mateu
Summary: In recent years, there have been sophisticated approaches proposed by authors to address road casualties and assist authorities in implementing new policies. These models usually consider socio-economic variables while ignoring the impact of measurement error on statistical inference. This paper presents a Bayesian model that analyzes car crash occurrences at the network-lattice level, accounting for measurement error in spatial covariates. The methodology is demonstrated using collision data from the road network in Leeds (UK) between 2011 and 2019, with traffic volumes approximated from extensive counts collected through mobile devices and adjusted using spatial measurement error correction.
JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES A-STATISTICS IN SOCIETY
(2023)
Article
Mathematics, Interdisciplinary Applications
Julia Calatayud, Marc Jornet, Jorge Mateu, Carla M. A. Pinto
Summary: This study investigates the infestation of rats and cockroaches in Madrid, Spain using differential equation models. Analyzing incidence and seasonal and weather factors is crucial for intervention strategies. The models can be used to predict future infestation dynamics, guiding health policy measures.
CHAOS SOLITONS & FRACTALS
(2023)
