Statistical Pattern Research Articles

The problem of alignment

AbstractLarge language models (LLMs) produce sequences learned as statistical patterns from large corpora. Their emergent status as representatives of the advances in artificial intelligence (AI) have led to an increased attention to the possibilities of regulating the automated production of linguistic utterances and interactions with human users in a process that computer scientists refer to as ‘alignment’—a series of technological and political mechanisms to impose a normative model of morality on algorithms and networks behind the model. Alignment, which can be viewed as the superimposition of normative structure onto a statistical model, however, reveals a conflicted and complex history of the conceptualisation of an interrelationship between language, mind and technology. This relationship is shaped by and, in turn, influences theories of language, linguistic practice and subjectivity, which are especially relevant to the current sophistication in artificially produced text. In this paper, we propose a critical evaluation of the concept of alignment, arguing that the theories and practice behind LLMs reveal a more complex social and technological dynamic of output coordination. We examine this dynamic as a two-way interaction between users and models by analysing how ChatGPT4 redacts perceived ‘anomalous’ language in fragments of Joyce’s Ulysses. We then situate this alignment problem historically, revisiting earlier postwar linguistic debates which counterposed two views of meaning: as discrete structures, and as continuous probability distributions. We discuss the largely occluded work of the Moscow Linguistic School, which sought to reconcile this opposition. Our attention to the Moscow School and later related arguments by Searle and Kristeva casts the problem of alignment in a new light: as one involving attention to the social regulation of linguistic practice, including rectification of anomalies that, like the Joycean text, exist in defiance of expressive conventions. The “problem of alignment” that we address here is, therefore, twofold: on one hand, it points to its narrow and normative definition in current technological development and critical research and, on the other hand, to the reality of complex and contradictory relations between subjectivity, technology and language that alignment problems reveal.

Daily sunshine grids for Austria since 1961 – combining station and satellite observations for a multi-decadal climate-monitoring dataset

Grid datasets of sunshine duration at high spatial resolution and extending over many decades are required for quantitative applications in regional climatology and environmental change (e.g., modelling of droughts and snow/ice covers, evaluation of clouds in numerical models, mapping of solar energy potentials). We present a new gridded dataset of relative (and derived absolute) sunshine duration for Austria at a grid spacing of 1 km, extending back until 1961 at daily time resolution. Challenges in the dataset construction were consistency issues in the available station data, the scarcity of long time series, and the high variation of cloudiness in the study region. The challenges were addressed by special efforts to correct evident breaks in the station series and by adopting an analysis method, which combines station data with satellite data. The methodology merges the data sources non-contemporaneously, using statistical patterns distilled over a short period, which allowed involving satellite data even for the early part of the study period. The resulting fields contain plausible mesoscale structures, which could not be resolved by the station network alone. On average, the analyses explain 47% of the spatial variance in daily sunshine duration at the stations. Evaluation revealed a slight systematic underestimation (− 1.5%) and a mean absolute error of 9.2%. The average error is larger during winter, at high altitudes, and around the 1990s. The dataset exhibits a conditional bias, which can lead to considerable systematic errors (up to 15%) when calculating sunshine-related climate indices.

Role of ChatGPT in antenatal care

ChatGPT, developed by OpenAI, represents a significant advancement in AI technology, particularly noted for its ability to engage in human-like conversations. This review article is about role of ChatGPT in antenatal care during pregnancy and its limitations in addressing health-related issues. While ChatGPT offers accessible and empathetic responses to queries about pregnancy and childbirth, its reliability and accuracy remain concerns. The model's reliance on statistical patterns in data often leads to biased or incomplete information, lacking necessary citations. Despite these limitations, ChatGPT shows promise in supporting clinical decision-making but requires oversight to ensure medical advice is precise and appropriate. Ethical considerations, including privacy and misuse risks, underscore the need for further development and validation in healthcare applications.

Promoting the Shift From Pixel-Level Correlations to Object Semantics Learning by Rethinking Computer Vision Benchmark Data Sets.

In computer vision research, convolutional neural networks (CNNs) have demonstrated remarkable capabilities at extracting patterns from raw pixel data, achieving state-of-the-art recognition accuracy. However, they significantly differ from human visual perception, prioritizing pixel-level correlations and statistical patterns, often overlooking object semantics. To explore this difference, we propose an approach that isolates core visual features crucial for human perception and object recognition: color, texture, and shape. In experiments on three benchmarks-Fruits 360, CIFAR-10, and Fashion MNIST-each visual feature is individually input into a neural network. Results reveal data set-dependent variations in classification accuracy, highlighting that deep learning models tend to learn pixel-level correlations instead of fundamental visual features. To validate this observation, we used various combinations of concatenated visual features as input for a neural network on the CIFAR-10 data set. CNNs excel at learning statistical patterns in images, achieving exceptional performance when training and test data share similar distributions. To substantiate this point, we trained a CNN on CIFAR-10 data set and evaluated its performance on the "dog" class from CIFAR-10 and on an equivalent number of examples from the Stanford Dogs data set. The CNN poor performance on Stanford Dogs images underlines the disparity between deep learning and human visual perception, highlighting the need for models that learn object semantics. Specialized benchmark data sets with controlled variations hold promise for aligning learned representations with human cognition in computer vision research.

Quantitative and Qualitative Analysis of Moral Foundations in Argumentation

This paper introduces moral argument analytics, a technology that provides insights into the use of moral arguments in discourse. We analyse five socio-political corpora of argument annotated data from offline and online discussions, totalling 240k words with 9k arguments, with an average annotation accuracy of 78%. Using a lexicon-based method, we automatically annotate these arguments with moral foundations, achieving an estimated accuracy of 83%. Quantitative analysis allows us to observe statistical patterns and trends in the use of moral arguments, whereas qualitative analysis enables us to understand and explain the communication strategies in the use of moral arguments in different settings. For instance, supporting arguments often rely on Loyalty and Authority, while attacking arguments use Care. We find that online discussions exhibit a greater diversity of moral foundations and a higher negative valence of moral arguments. Online arguers often rely more on Harm rather than Care, Degradation rather than Sanctity. These insights have significant implications for AI applications, particularly in understanding and predicting human and machine moral behaviours. This work contributes to the construction of more convincing messages and the detection of harmful or biased AI-generated synthetic content.

Productive explanation: A framework for evaluating explanations in psychological science.

The explanation of psychological phenomena is a central aim of psychological science. However, the nature of explanation and the processes by which we evaluate whether a theory explains a phenomenon are often unclear. Consequently, it is often unknown whether a given psychological theory indeed explains a phenomenon. We address this shortcoming by proposing a productive account of explanation: a theory explains a phenomenon to some degree if and only if a formal model of the theory produces the statistical pattern representing the phenomenon. Using this account, we outline a workable methodology of explanation: (a) explicating a verbal theory into a formal model, (b) representing phenomena as statistical patterns in data, and (c) assessing whether the formal model produces these statistical patterns. In addition, we provide three major criteria for evaluating the goodness of an explanation (precision, robustness, and empirical relevance), and examine some cases of explanatory breakdowns. Finally, we situate our framework within existing theories of explanation from philosophy of science and discuss how our approach contributes to constructing and developing better psychological theories. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Safety Evaluation for Fabricated Small Box Girder Bridges Based on Fuzzy Analytic Hierarchy Process and Monitoring Data.

During the operation of fabricated small box girder bridges, which face safety issues such as structural degradation and failure, there is an urgent need to propose a safety evaluation method to cope with the possible risks. This article quantitatively evaluates the safety state of a fabricated small box girder bridge in Wuhan City based on Fuzzy Analytic Hierarchy Process (FAHP) and structural health monitoring (SHM) data. Firstly, the FAHP model is established, and stress, deformation, and temperature are selected as evaluation factors. The safety thresholds of stress and deformation are determined by combining the industry specifications and the historical statistical patterns of the massive SHM data. The temperature field of the bridge is simulated and analyzed by combining ANSYS, HYPERMESH, and TAITHREM, and the most unfavorable temperature gradient is determined as a threshold for the safety evaluation. Finally, the scores of indexes of the bridge are determined based on the measured SHM data, which in turn provides a quantitative description of the safety state. The results show that the thresholds determined by the joint industry specifications and the massive SHM data are reasonable; the temperature field simulation model established in this article is consistent with the measured results, and can accurately determine the temperature gradient of the bridge. The safety evaluation result from the FAHP model is the same as the field test results, which verifies the effectiveness and applicability of the proposed method to actual bridge projects.

Causal Machine Learning in Marketing

This study reviews three primary purposes of causal machine learning (CML) in marketing, merging impact evaluation of marketing interventions with machine learning algorithms for learning statistical patterns from data. Firstly, CML enables more credible impact evaluation by considering important control variables that simultaneously influence the intervention and business outcomes (such as sales) in a data-driven manner. Secondly, it facilitates the data-driven detection of customer segments for which a marketing intervention is particularly effective or ineffective, a process known as effect moderation or heterogeneity analysis. Thirdly, closely related to the second point, it allows for optimal customer segmentation into groups that should and should not be targeted by the intervention to maximize overall effectiveness. The discussion is grounded in recent empirical applications, all of which aim to enhance decision support in marketing by leveraging data-driven evaluation and optimization of interventions across different customer groups.

New mechanistic understanding of osteoclast differentiation and bone resorption mediated by P2X7 receptors and PI3K-Akt-GSK3β signaling

ObjectiveOsteoporosis is a global health issue characterized by decreased bone mass and microstructural degradation, leading to an increased risk of fractures. This study aims to explore the molecular mechanism by which P2X7 receptors influence osteoclast formation and bone resorption through the PI3K-Akt-GSK3β signaling pathway.MethodsAn osteoporosis mouse model was generated through ovariectomy (OVX) in normal C57BL/6 and P2X7f/f; LysM-cre mice. Osteoclasts were isolated for transcriptomic analysis, and differentially expressed genes were selected for functional enrichment analysis. Metabolite analysis was performed using liquid chromatography-tandem mass spectrometry (LC–MS/MS), and multivariate statistical analysis and pattern recognition were used to identify differential lipid metabolism markers and their distribution. Bioinformatics analyses were conducted using the Encyclopedia of Genes and Genomes database and the MetaboAnalyst database to assess potential biomarkers and create a metabolic pathway map. Osteoclast precursor cells were used for in vitro cell experiments, evaluating cell viability and proliferation using the Cell Counting Kit 8 (CCK-8) assay. Osteoclast precursor cells were induced to differentiate into osteoclasts using macrophage colony-stimulating factor (M-CSF) and receptor activator of nuclear factor kappa-beta ligand (RANKL), and tartrate-resistant acid phosphatase (TRAP) staining was performed to compare differentiation morphology, size, and quantity between different groups. Western blot analysis was used to assess the expression of differentiation markers, fusion gene markers, and bone resorption ability markers in osteoclasts. Immunofluorescence staining was employed to examine the spatial distribution and quantity of osteoclast cell skeletons, P2X7 protein, and cell nuclei, while pit assay was used to evaluate osteoclast bone resorption ability. Finally, in vivo animal experiments, including micro computed tomography (micro-CT), hematoxylin and eosin (HE) staining, TRAP staining, and immunohistochemistry, were conducted to observe bone tissue morphology, osteoclast differentiation, and the phosphorylation level of the PI3K-Akt-GSK3β signaling pathway.ResultsTranscriptomic and metabolomic data collectively reveal that the P2X7 receptor can impact the pathogenesis of osteoporosis through the PI3K-Akt-GSK3β signaling pathway. Subsequent in vitro experiments showed that cells in the Sh-P2X7 + Recilisib group exhibited increased proliferative activity (1.15 versus 0.59), higher absorbance levels (0.68 versus 0.34), and a significant increase in resorption pit area (13.94 versus 3.50). Expression levels of osteoclast differentiation-related proteins MMP-9, CK, and NFATc1 were markedly elevated (MMP-9: 1.72 versus 0.96; CK: 2.54 versus 0.95; NFATc1: 3.05 versus 0.95), along with increased fluorescent intensity of F-actin rings. In contrast, the OE-P2X7 + LY294002 group showed decreased proliferative activity (0.64 versus 1.29), reduced absorbance (0.34 versus 0.82), and a significant decrease in resorption pit area (5.01 versus 14.96), accompanied by weakened expression of MMP-9, CK, and NFATc1 (MMP-9: 1.14 versus 1.79; CK: 1.26 versus 2.75; NFATc1: 1.17 versus 2.90) and decreased F-actin fluorescent intensity. Furthermore, in vivo animal experiments demonstrated that compared with the wild type (WT) + Sham group, mice in the WT + OVX group exhibited significantly increased levels of CTX and NTX in serum (CTX: 587.17 versus 129.33; NTX: 386.00 versus 98.83), a notable decrease in calcium deposition (19.67 versus 53.83), significant reduction in bone density, increased trabecular separation, and lowered bone mineral density (BMD). When compared with the KO + OVX group, mice in the KO + OVX + recilisib group showed a substantial increase in CTX and NTX levels in serum (CTX: 503.50 versus 209.83; NTX: 339.83 versus 127.00), further reduction in calcium deposition (29.67 versus 45.33), as well as decreased bone density, increased trabecular separation, and reduced BMD.ConclusionP2X7 receptors positively regulate osteoclast formation and bone resorption by activating the PI3K-Akt-GSK3β signaling pathway.

Description of regression models for predicting the dynamics of pink salmon returns in the Kamchatka region based on climate-oceanological and population-genetic data

Several regression models for predicting returns of pink salmon in the Kamchatka region are presented. The data for 1990–2023 were analyzed. Among available climatic and oceanological indices, the most suitable for using as predictors for forecasting of pink salmon returns were the Pacific Decadal Oscillation (PDO) index, Western Pacific Cyclonic Index (WP), Arctic Oscillation (AO) index, and the sea surface temperature anomaly in the North Pacific. Multi-dimensional models of the «stock–recruitment» type were built on identified statistical patterns, which allowed to estimate potential abundance of the pink salmon returns to northeastern and western Kamchatka. Besides, methods for predicting the abundance of pink salmon returns on the data of fish counting in the sea are considered, using the materials of TINRO trawl surveys conducted in the Bering and Okhotsk Seas in the fall seasons of 2012–2023. To determine the abundance of pink salmon originated from West Kamchatka, genetic identification of regional composition of juveniles in mixed trawl catches was used. All tested methods have a high level of determination, but simpler regressive models are more prospective for practical forecasting of general trend in dynamics of pink salmon stocks in the Kamchatka region due to very weak generalization ability of more complicated models.

СУИЦИДАЛЬНАЯ АКТИВНОСТЬ В СТАТИСТИЧЕСКОЙ РЕТРОСПЕКЦИИ

Purpose. To determine the statistical features of the frequency of suicides in the Grodno region with regard to time intervals as well as medical and social risk factors for suicide. Material and methods. In the present study we used statistical data for the Grodno region for the period 2004-2023 concerning the population size and the number of suicides as well as the maps of analysis of the causes of suicides committed in 2010-2023 (n=1520). Results. The established statistical features of the frequency of suicides in the Grodno region indicate that suicidal activity, as a population feature, is independent of the population size; it is determined and has characteristic dynamics over annual intervals and extremes within the year, month, week and day. In the structure of pre-suicide, medical and social factors not related to health conditions and comorbid with alcohol abuse predominate. Сonclusions. The established in the study statistical patterns of suicidal activity of the population of the Grodno region, which are quantitatively associated with time intervals within the day, week, month and year and qualitatively with the dominant medical and social characteristics present in pre-suicide, indicate that suicide is a population feature and causes of suicides in particular cases are determined by the suicider’s inheriting of some features of a society as an element and a formal model of this society.

Coexistence of Standard Modifiable, Other Classical, and Novel and Classical Atherosclerotic Cardiovascular Disease Risk Factors in Middle Eastern Young Women.

The coexistence of multiple standard modifiable risk factors (SMuRFs),classical and novel risk factors (RFs) for atherosclerotic cardiovascular disease (ASCVD) is common in the Middle East (ME). There is a paucity of data on the coexistence of these RFs in ME young women. Comparing the prevalence and the statistical patterns of the SMuRFs, classical and novel RFs in target population. In this case-control (1:2) study, consecutive young women aged 18-50 years were enrolled in 12 centers (July 2021 to October 2023). Prevalence and coexistence of 19 RFs were compared between cases with ASCVD and their controls. The RFs included SMuRFs (hypertension, type 2 diabetes, dyslipidemia, and cigarette smoking), other classical RF (obesity, family history of premature ASCVD, and physical inactivity), novel RFs and social determinants of health (health insurance, place of residence, depression, and level of education). The study included 627 subjects; 209 had ASCVD (median age 46 years, IQR 49-42 years) and 418 controls (median age 45 years, IQR 48-41 years). The presence of 1-2 RFs; (ASCVD: 63.2%, Control: 54.1%, p=0.037) and 3-4 RFs; (ASCVD: 27.8%, Control: 3.3%, p < 0.001) SMuRFs was more prevalent in women with ASCVD. Similarly, the presence of 4-5 RFs; (ASCVD: 40.7%, Control: 14.6%, p<0.001), and 6-7 (ASCVD: 10.5%, Control: 1%, p < 0.001). The classical RF were also significantly common in these women. The distribution of multiple novel RF was not statistically significant across both groups. Finally, regarding the socioeconomic RFs in women with ASCVDs, the presence of 1-2 RFs (ASCVD: 59.8%, Control: 76.1%, p < 0.001) was significantly less common while the presence of 3-4 RFs (ASCVD: 39.2%, Control: 21.8%, p < 0.001) was vastly more common. An elevated rate of coexistence of classical RF in the case group, mainly socioeconomic and SMuRFs. By managing them primary and secondary ASCVDs prevention attained.

Spatial patterns of human–long-tailed macaque (Macaca fascicularis) conflicts in Java Island: A comparison of two secondary data sources

Human–macaque conflicts present a significant obstacle to conservation and management efforts in Indonesia. Long-tailed macaques, a species increasingly in conflict with humans, are often regarded as pests due to behaviors such as occasional raiding on agricultural land. This study aims to offer comprehensive insights into these conflicts by comparing data from two different sources and analyzing their spatial patterns to identify regions with significant conflict occurrences. Official records show a larger quantity of data, while online news supplements this. However, official records data quality varies, with diverse data formats and collection methods, whereas online news data consistently provide accessible information over a longer period. Both sources exhibit similar global spatial statistical patterns, indicating significant clustering. Local autocorrelation analysis identifies hotspot areas primarily at the borders of Yogyakarta, Central Java, and East Java. Discrepancies between the sources are evident in the High-Low class and the lack of clustering patterns in the western part of Java Island in official records. Despite similar spatial patterns, official records do not fully reflect the actual human–macaque conflict. Standardized methods and interview surveys are crucial for accurately assessing this conflict. Citizen science offers the potential for long-term wildlife monitoring, requiring systematic collaboration among stakeholders to mobilize experienced volunteers.

Enhancing Roads Infrastructure Monitoring with FBG Sensors and Multi-Threshold Signal Tracking

Accurate monitoring techniques are needed to improve pavement durability by ensuring their structural integrity and longevity. This study developed an innovative real-time monitoring system utilizing fiber Bragg grating (FBG) sensors embedded in asphalt layers. The data collected by FBG sensors includes a baseline primarily influenced by temperature and peak shape data associated with short-term loading from passing vehicles. Therefore, analyzing FBG data and identifying statistical patterns poses a challenge. To address this challenge, employing a signal tracker proves to be a valuable option for monitoring the baseline and scanning signals for subsequent analysis. Moreover, considering the varying loading weights on FBG sensors due to diverse vehicles, tracking FBG signals across multiple thresholds enhances precision. A MATLAB-based program was developed incorporating a multi-threshold signal tracker capable of efficiently processing FBG data by tracking signals in both positive and negative multi-thresholds. This tracker effectively distinguishes between different loading events and extracts the baseline signal after tracking. Subsequently, the system provides outputs, including loading distribution, number of loading events, traffic, and temperature effects. Finally, to validate the system, real FBG data collected during an extensive monitoring campaign was processed using the developed program. The obtained outputs contribute to a comprehensive assessment of pavement performance, thereby facilitating the development of more accurate maintenance and rehabilitation strategies.

Probability ogives for trends in stock biomass and fishing mortality from landings time series

AbstractMost fisheries are conducted without any scientific knowledge about the size and productivity of the stocks that support them. This navigation in the dark in most fisheries is a major obstacle in making them sustainable sources of nutrition for people in general and income for fishers and other economic actors along supply chains. Fisheries that have not been assessed generally are data‐intermediate and data‐poor, the latter usually having annual time series of landings as the single piece of data available. A major effort in the last two decades has been directed toward developing ‘catch‐only’ stock assessment methods, although some of these methods have been tested and found deficient. Here we provide a novel approach to using annual landing time series as the single source of data to qualitatively judge the condition of un‐assessed stocks using frequentist cumulative probability ogives, both in terms of stock biomass and fishing mortality. A meta‐analysis of the FishSource database allowed us to infer statistical patterns from hundreds of assessed fisheries and thousands of annual landings, biomass, and fishing mortality observations. Four stock‐management types were considered separately in the analysis: short‐lived and others (mid‐ to long‐lived) stocks, controlled or not controlled by catch limits. Obtained cumulative probability ogives provide clear evaluations of stock biomass and fishing mortality trends in all four stock‐management types, leading to actionable information on probable current status and future trends. Using these probability ogives, we developed decision trees that lead to qualitative scores on the exploitation status of un‐assessed stocks.

Three-dimensional characterization of abrasive chips using micro-computed tomography

Chip formation is considered one important indicator to evaluate machining processes. In particular, geometric features of chips may provide important insights for the assessment of machining stability and productivity. In fixed-abrasive machining processes, such as grinding and honing, chips are simultaneously produced by many of the geometrically undefined cutting edges. Despite being “undefined,” geometric features of abrasive grains can still be statistically characterized or described. Accordingly, it can be assumed that, under stable machining conditions, the geometric features of abrasive chips may also conform to certain statistical patterns. However, statistical characterization of abrasive chips can be very challenging due to their large quantity, irregular shapes, minuscule size, and sometimes tangled condition. In this study, an analysis method combining metallographic preparation and micro-computed tomography (micro-CT) has been developed to characterize abrasive chips produced by a honing stone. The results regarding the geometric characteristics of the obtained massive abrasive chips, i.e., their sizes and shapes, were presented and statistically described. It was found most chips had a size around 50 µm and the shape being a slightly elongated and curved cone. Most of the geometric features could meet a positive skewness distribution.

Data-driven fine-grained region discovery in the mouse brain with transformers.

Technologies such as spatial transcriptomics offer unique opportunities to define the spatial organization of the mouse brain. We developed an unsupervised training scheme and novel transformer-based deep learning architecture to detect spatial domains across the whole mouse brain using spatial transcriptomics data. Our model learns local representations of molecular and cellular statistical patterns which can be clustered to identify spatial domains within the brain from coarse to fine-grained. Discovered domains are spatially regular, even with several hundreds of spatial clusters. They are also consistent with existing anatomical ontologies such as the Allen Mouse Brain Common Coordinate Framework version 3 (CCFv3) and can be visually interpreted at the cell type or transcript level. We demonstrate our method can be used to identify previously uncatalogued subregions, such as in the midbrain, where we uncover gradients of inhibitory neuron complexity and abundance. Notably, these subregions cannot be discovered using other methods. We apply our method to a separate multi-animal whole-brain spatial transcriptomic dataset and show that our method can also robustly integrate spatial domains across animals.

Prevalence of gastritis and gastric and duodenal ulcers in Russian Guard servicemen

Based on the annual statistical reports for 2018–2023 and the cumulative median prevalence of various studies in 2009–2020 in Russia, the prevalence of peptic ulcer with uncomplicated and complicated course and various forms of gastritis in servicemen of the Regardie of the North Caucasian District was estimated, and the current trends are explained. The incidence of peptic ulcer with uncomplicated and complicated course and various forms of gastritis among the servicemen of the Regardie of the North Caucasian District for the specified period ranged from 11.521% to 16.351% (confidence interval: 10.515–12.588 and 15.233–17.517, respectively) and from 27.212% to 51.889% (confidence interval: 25.791–28.669 and 50.316–53.460, respectively). Random effects analysis has demonstrated different true effect sizes relative to the overall average. The revealed statistical patterns are only the first stage in an ongoing study on the prevalence of upper gastrointestinal diseases among the servicemen of the Regardie of the North Caucasian District. In general, the prevalence of peptic ulcer disease, including with a complicated course, and various forms of gastritis among military personnel was 13.270% and 37.679%, respectively (confidence interval: 10.617–16.170 and 24.292–52.102, respectively), showing an upward trend. The expected prevalence of peptic ulcer and gastritis in 2023 was 16,351 and 51,889%, respectively, which is not accidental and depends on the impact of specific risk factors. The identification of risk factors enables the development of measures to prevent delayed complications of the pathology under study and improvement of health in servicemen of the Regardie of the North Caucasian District. Thus, the study of the health status of the general population and military personnel, in particular, and the effect of various environmental factors determines their social significance both in military medicine and the structure of the healthcare system.

Dilemma of the investment strategy: dividend cut-offs (retrospective analysis)

The author considers the results of a statistical analysis of changes in stock prices for 10 largest Russian companies in various industries over 12 years in connection with dividend payments. Statistical patterns of stock price behavior before and after the dividend cut-offs are revealed, which allows investors to build their dividend strategies and answer the “eternal” question of question of whether to sell or to buy shares before the dividend cut-off. A retrospective statistical analysis is carried out on 120 dividend payments.

Early Warning for Continuous Rigid Frame Bridges Based on Nonlinear Modeling for Temperature-Induced Deflection.

Bridge early warning based on structural health monitoring (SHM) system is of significant importance for ensuring bridge safe operation. The temperature-induced deflection (TID) is a sensitive indicator for performance degradation of continuous rigid frame bridges, but the time-lag effect makes it challenging to predict the TID accurately. A bridge early warning method based on nonlinear modeling for the TID is proposed in this article. Firstly, the SHM data of temperature and deflection of a continuous rigid frame bridge are analyzed to examine the temperature gradient variation patterns. Kernel principal component analysis (KPCA) is used to extract principal temperature components. Then, the TID is extracted through wavelet transform, and a nonlinear modeling method for the TID considering the temperature gradient is proposed using the support vector machine (SVM). Finally, the prediction errors of the KPCA-SVM algorithm are analyzed, and the early warning thresholds are determined based on the statistical patterns of the errors. The results show that the KPCA-SVM algorithm achieves high-precision nonlinear modeling for the TID while significantly reducing the computational load. The prediction results have coefficients of determination above 0.98 and fluctuate within a small range with clear statistical patterns. Setting the early warning thresholds based on the statistical patterns of errors enables dynamic and multi-level warnings for bridge structures.