arXiv - STAT - Applications最新文献

{"title":"Spatial occupancy models for data collected on stream networks","authors":"Olivier Gimenez","doi":"arxiv-2409.10017","DOIUrl":"https://doi.org/arxiv-2409.10017","url":null,"abstract":"To effectively monitor biodiversity in streams and rivers, we need to\u0000quantify species distribution accurately. Occupancy models are useful for\u0000distinguishing between the non-detection of a species and its actual absence.\u0000While these models can account for spatial autocorrelation, they are not suited\u0000for streams and rivers due to their unique network spatial structure. Here, I\u0000propose spatial occupancy models specifically designed for data collected on\u0000stream and river networks. I present the statistical developments and\u0000illustrate their application using data on a semi-aquatic mammal. Overall,\u0000spatial stream network occupancy models offer a robust method for assessing\u0000biodiversity in freshwater ecosystems.","PeriodicalId":501172,"journal":{"name":"arXiv - STAT - Applications","volume":"53 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142261492","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Leadership and Engagement Dynamics in Legislative Twitter Networks: Statistical Analysis and Modeling","authors":"Carolina Luque, Juan Sosa","doi":"arxiv-2409.10475","DOIUrl":"https://doi.org/arxiv-2409.10475","url":null,"abstract":"In this manuscript, we analyze the interaction network on Twitter among\u0000members of the 117th U.S. Congress to assess the visibility of political\u0000leaders and explore how systemic properties and node attributes influence the\u0000formation of legislative connections. We employ descriptive social network\u0000statistical methods, the exponential random graph model (ERGM), and the\u0000stochastic block model (SBM) to evaluate the relative impact of network\u0000systemic properties, as well as institutional and personal traits, on the\u0000generation of online relationships among legislators. Our findings reveal that\u0000legislative networks on social media platforms like Twitter tend to reinforce\u0000the leadership of dominant political actors rather than diminishing their\u0000influence. However, we identify that these leadership roles can manifest in\u0000various forms. Additionally, we highlight that online connections within\u0000legislative networks are influenced by both the systemic properties of the\u0000network and institutional characteristics.","PeriodicalId":501172,"journal":{"name":"arXiv - STAT - Applications","volume":"49 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142261490","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Convolutional Neural Network-based Ensemble Post-processing with Data Augmentation for Tropical Cyclone Precipitation Forecasts","authors":"Sing-Wen ChenInstitute of Health Data Analytics and Statistics, College of Public Health, National Taiwan University, Taiwan, Joyce JuangCentral Weather Administration, Taiwan, Charlotte WangInstitute of Health Data Analytics and Statistics, College of Public Health, National Taiwan University, TaiwanMaster Program of Public Health, College of Public Health, National Taiwan University, Taiwan, Hui-Ling ChangCentral Weather Administration, Taiwan, Jing-Shan HongCentral Weather Administration, Taiwan, Chuhsing Kate HsiaoInstitute of Health Data Analytics and Statistics, College of Public Health, National Taiwan University, TaiwanMaster Program of Public Health, College of Public Health, National Taiwan University, Taiwan","doi":"arxiv-2409.09607","DOIUrl":"https://doi.org/arxiv-2409.09607","url":null,"abstract":"Heavy precipitation from tropical cyclones (TCs) may result in disasters,\u0000such as floods and landslides, leading to substantial economic damage and loss\u0000of life. Prediction of TC precipitation based on ensemble post-processing\u0000procedures using machine learning (ML) approaches has received considerable\u0000attention for its flexibility in modeling and its computational power in\u0000managing complex models. However, when applying ML techniques to TC\u0000precipitation for a specific area, the available observation data are typically\u0000insufficient for comprehensive training, validation, and testing of the ML\u0000model, primarily due to the rapid movement of TCs. We propose to use the\u0000convolutional neural network (CNN) as a deep ML model to leverage the spatial\u0000information of precipitation. The proposed model has three distinct features\u0000that differentiate it from traditional CNNs applied in meteorology. First, it\u0000utilizes data augmentation to alleviate challenges posed by the small sample\u0000size. Second, it contains geographical and dynamic variables to account for\u0000area-specific features and the relative distance between the study area and the\u0000moving TC. Third, it applies unequal weights to accommodate the temporal\u0000structure in the training data when calculating the objective function. The\u0000proposed CNN-all model is then illustrated with the TC Soudelor's impact on\u0000Taiwan. Soudelor was the strongest TC of the 2015 Pacific typhoon season. The\u0000results show that the inclusion of augmented data and dynamic variables\u0000improves the prediction of heavy precipitation. The proposed CNN-all\u0000outperforms traditional CNN models, based on the continuous probability skill\u0000score (CRPSS), probability plots, and reliability diagram. The proposed model\u0000has the potential to be utilized in a wide range of meteorological studies.","PeriodicalId":501172,"journal":{"name":"arXiv - STAT - Applications","volume":"72 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142261494","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Zipf's law in the distribution of Brazilian firm size","authors":"Thiago Trafane Oliveira SantosCentral Bank of Brazil, Brasília, Brazil. Department of %Economics, University of Brasilia, Brazil, Daniel Oliveira CajueiroDepartment of Economics, University of Brasilia, Brazil. National Institute of Science and Technology for Complex Systems","doi":"arxiv-2409.09470","DOIUrl":"https://doi.org/arxiv-2409.09470","url":null,"abstract":"Zipf's law states that the probability of a variable being larger than $s$ is\u0000roughly inversely proportional to $s$. In this paper, we evaluate Zipf's law\u0000for the distribution of firm size by the number of employees in Brazil. We use\u0000publicly available binned annual data from the Central Register of Enterprises\u0000(CEMPRE), which is held by the Brazilian Institute of Geography and Statistics\u0000(IBGE) and covers all formal organizations. Remarkably, we find that Zipf's law\u0000provides a very good, although not perfect, approximation to data for each year\u0000between 1996 and 2020 at the economy-wide level and also for agriculture,\u0000industry, and services alone. However, a lognormal distribution also performs\u0000well and even outperforms Zipf's law in certain cases.","PeriodicalId":501172,"journal":{"name":"arXiv - STAT - Applications","volume":"85 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142261496","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Forensically useful mid-term and short-term temperature reconstruction for quasi-indoor death scenes","authors":"Jędrzej Wydra, Łukasz Smaga, Szymon Matuszewski","doi":"arxiv-2409.09516","DOIUrl":"https://doi.org/arxiv-2409.09516","url":null,"abstract":"Accurate reconstruction of ambient temperature at death scenes is crucial for\u0000estimating the postmortem interval (PMI) in forensic science. Typically, this\u0000is done by correcting weather station temperatures using measurements from the\u0000scene, often through linear regression. While recent attempts to use\u0000alternative algorithms like GAM have improved accuracy, they usually require\u0000additional variables such as humidity, making them impractical. This study\u0000presents two methods for accurate temperature reconstruction using only\u0000temperature data. The first, a concurrent regression model, is known in\u0000mathematics and is applied here for mid-term reconstructions (several days of\u0000measurements). The second, a new method based on Fourier expansion, is designed\u0000for short-term reconstructions (only a few hours of measurements). Both models\u0000were tested in quasi-indoor conditions, using data from six different\u0000environments. The concurrent regression model provided nearly perfect\u0000reconstructions for periods longer than six days, while the short-term model\u0000achieved similar accuracy after just 4-5 hours of measurements. These findings\u0000demonstrate that reliable temperature corrections for PMI estimation can be\u0000made with significantly reduced measurement periods, enhancing the practicality\u0000of the method in forensic applications.","PeriodicalId":501172,"journal":{"name":"arXiv - STAT - Applications","volume":"8 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142261495","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Exploring Dimensionality Reduction of SDSS Spectral Abundances","authors":"Qianyu Fan, Joshua S. Speagle","doi":"arxiv-2409.09227","DOIUrl":"https://doi.org/arxiv-2409.09227","url":null,"abstract":"High-resolution stellar spectra offer valuable insights into atmospheric\u0000parameters and chemical compositions. However, their inherent complexity and\u0000high-dimensionality present challenges in fully utilizing the information they\u0000contain. In this study, we utilize data from the Apache Point Observatory\u0000Galactic Evolution Experiment (APOGEE) within the Sloan Digital Sky Survey IV\u0000(SDSS-IV) to explore latent representations of chemical abundances by applying\u0000five dimensionality reduction techniques: PCA, t-SNE, UMAP, Autoencoder, and\u0000VAE. Through this exploration, we evaluate the preservation of information and\u0000compare reconstructed outputs with the original 19 chemical abundance data. Our\u0000findings reveal a performance ranking of PCA < UMAP < t-SNE < VAE <\u0000Autoencoder, through comparing their explained variance under optimized MSE.\u0000The performance of non-linear (Autoencoder and VAE) algorithms has\u0000approximately 10% improvement compared to linear (PCA) algorithm. This\u0000difference can be referred to as the \"non-linearity gap.\" Future work should\u0000focus on incorporating measurement errors into extension VAEs, thereby\u0000enhancing the reliability and interpretability of chemical abundance\u0000exploration in astronomical spectra.","PeriodicalId":501172,"journal":{"name":"arXiv - STAT - Applications","volume":"23 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142261150","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"AutoIRT: Calibrating Item Response Theory Models with Automated Machine Learning","authors":"James Sharpnack, Phoebe Mulcaire, Klinton Bicknell, Geoff LaFlair, Kevin Yancey","doi":"arxiv-2409.08823","DOIUrl":"https://doi.org/arxiv-2409.08823","url":null,"abstract":"Item response theory (IRT) is a class of interpretable factor models that are\u0000widely used in computerized adaptive tests (CATs), such as language proficiency\u0000tests. Traditionally, these are fit using parametric mixed effects models on\u0000the probability of a test taker getting the correct answer to a test item\u0000(i.e., question). Neural net extensions of these models, such as BertIRT,\u0000require specialized architectures and parameter tuning. We propose a multistage\u0000fitting procedure that is compatible with out-of-the-box Automated Machine\u0000Learning (AutoML) tools. It is based on a Monte Carlo EM (MCEM) outer loop with\u0000a two stage inner loop, which trains a non-parametric AutoML grade model using\u0000item features followed by an item specific parametric model. This greatly\u0000accelerates the modeling workflow for scoring tests. We demonstrate its\u0000effectiveness by applying it to the Duolingo English Test, a high stakes,\u0000online English proficiency test. We show that the resulting model is typically\u0000more well calibrated, gets better predictive performance, and more accurate\u0000scores than existing methods (non-explanatory IRT models and explanatory IRT\u0000models like BERT-IRT). Along the way, we provide a brief survey of machine\u0000learning methods for calibration of item parameters for CATs.","PeriodicalId":501172,"journal":{"name":"arXiv - STAT - Applications","volume":"17 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142261149","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Cross-Country Comparative Analysis of Climate Resilience and Localized Mapping in Data-Sparse Regions","authors":"Ronald Katende","doi":"arxiv-2409.08765","DOIUrl":"https://doi.org/arxiv-2409.08765","url":null,"abstract":"Climate resilience across sectors varies significantly in low-income\u0000countries (LICs), with agriculture being the most vulnerable to climate change.\u0000Existing studies typically focus on individual countries, offering limited\u0000insights into broader cross-country patterns of adaptation and vulnerability.\u0000This paper addresses these gaps by introducing a framework for cross-country\u0000comparative analysis of sectoral climate resilience using meta-analysis and\u0000cross-country panel data techniques. The study identifies shared\u0000vulnerabilities and adaptation strategies across LICs, enabling more effective\u0000policy design. Additionally, a novel localized climate-agriculture mapping\u0000technique is developed, integrating sparse agricultural data with\u0000high-resolution satellite imagery to generate fine-grained maps of agricultural\u0000productivity under climate stress. Spatial interpolation methods, such as\u0000kriging, are used to address data gaps, providing detailed insights into\u0000regional agricultural productivity and resilience. The findings offer\u0000policymakers tools to prioritize climate adaptation efforts and optimize\u0000resource allocation both regionally and nationally.","PeriodicalId":501172,"journal":{"name":"arXiv - STAT - Applications","volume":"16 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142261151","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Statistical Analysis of Quantitative Cancer Imaging Data","authors":"Shariq Mohammed, Maria Masotti, Nathaniel Osher, Satwik Acharyya, Veerabhadran Baladandayuthapani","doi":"arxiv-2409.08809","DOIUrl":"https://doi.org/arxiv-2409.08809","url":null,"abstract":"Recent advances in types and extent of medical imaging technologies has led\u0000to proliferation of multimodal quantitative imaging data in cancer.\u0000Quantitative medical imaging data refer to numerical representations derived\u0000from medical imaging technologies, such as radiology and pathology imaging,\u0000that can be used to assess and quantify characteristics of diseases, especially\u0000cancer. The use of such data in both clinical and research setting enables\u0000precise quantifications and analyses of tumor characteristics that can\u0000facilitate objective evaluation of disease progression, response to therapy,\u0000and prognosis. The scale and size of these imaging biomarkers is vast and\u0000presents several analytical and computational challenges that range from\u0000high-dimensionality to complex structural correlation patterns. In this review\u0000article, we summarize some state-of-the-art statistical methods developed for\u0000quantitative medical imaging data ranging from topological, functional and\u0000shape data analyses to spatial process models. We delve into common imaging\u0000biomarkers with a focus on radiology and pathology imaging in cancer, address\u0000the analytical questions and challenges they present, and highlight the\u0000innovative statistical and machine learning models that have been developed to\u0000answer relevant scientific and clinical questions. We also outline some\u0000emerging and open problems in this area for future explorations.","PeriodicalId":501172,"journal":{"name":"arXiv - STAT - Applications","volume":"19 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142261500","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Bayesian framework to evaluate evidence in cases of alleged cheating with secret codes in sports","authors":"Aafko Boonstra, Ronald Meester","doi":"arxiv-2409.08172","DOIUrl":"https://doi.org/arxiv-2409.08172","url":null,"abstract":"We present a Bayesian framework to analyze a case of alleged cheating in the\u0000mind sport contract bridge. We explain why a Bayesian approach is called for,\u0000and not a frequentistic one. We argue that such a Bayesian framework can and\u0000should also be used in other sports for cases of alleged cheating by means of\u0000illegal signalling.","PeriodicalId":501172,"journal":{"name":"arXiv - STAT - Applications","volume":"41 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142224680","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}