Class Model Research Articles

Soil textural class modeling using digital soil mapping approaches: Effect of resampling strategies on imbalanced dataset predictions

In a digital soil mapping (DSM) context, machine learning (ML) algorithms are widely used to model soil textural classes (STCs). However, in the real world most soil class datasets exhibit imbalanced distributions. This poses a challenge as ML algorithms are designed to handle balanced classes, leading to a bias towards the majority classes while often overlooking the minority classes. Furthermore, within the DSM framework, two strategies can be employed to model STCs: direct and indirect approaches. In the direct approach, STCs are directly inputted into the model for prediction. In contrast, the indirect approach involves introducing soil texture fractions (i.e., clay, silt, sand) as initial inputs, then STCs are obtained from the outputs. Limited research has been conducted on the impact of data balancing on STC predictions, and there is a lack of comparative analysis between direct and indirect approaches in this context. Therefore, this study aimed to evaluate the efficacy of a resampling technique (SMOTE: synthetic minority oversampling technique) in handling an imbalanced soil texture dataset collected from the Kuhdasht region in western Iran. Additionally, the study sought to compare the performance of direct and indirect modeling approaches. Environmental covariates derived from Landsat 8 and Sentinel 2 images along with a digital elevation model (DEM) were used as input variables to a random forest (RF) model to model STCs and soil texture fractions. The results revealed that terrain attributes and Euclidean distances played a more significant role in modeling both balanced and imbalanced datasets compared to remotely sensed data. Kappa indices for balanced and imbalanced datasets, as well as for the indirect approach were found to be 89%, 68% and 38% respectively. In the same way, the overall accuracies were 91%, 79% and 68%, respectively. Among the imbalanced classes, clay loam and loam which accounted for the majority of observations showed the highest recall values, followed by sandy clay loam, sandy loam and silty clay loam. When employing the indirect approach, the RF model failed to capture the minority classes in terms of validation statistics. Additionally, modeling with the imbalanced dataset resulted in the exclusion of three minority STCs from the final map. Overall, this study showed the importance of balancing STCs prior to modeling to achieve more accurate estimates of STCs, as well as the superiority of employing the direct approach (using balanced data sets) over the indirect approach.

Pediatric urinary tract infections caused by poultry-associated Escherichia coli.

Escherichia coli is the leading cause of urinary tract infections (UTIs) in children and adults. The gastrointestinal tract is the primary reservoir of uropathogenic E. coli, which can be acquired from a variety of environmental exposures, including retail meat. In the current study, we used a novel statistical-genomic approach to estimate the proportion of pediatric UTIs caused by foodborne zoonotic E. coli strains. E. coli urine isolates were collected from DC residents aged 2 months to 17 years from the Children's National Medical Center Laboratory, 2013-2014. During the same period, E. coli isolates were collected from retail poultry products purchased from 15 sites throughout DC. A total of 52 urine and 56 poultry isolates underwent whole-genome sequencing, core genome phylogenetic analysis, and host-origin prediction by a Bayesian latent class model that incorporated data on the presence of mobile genetic elements (MGEs) among E. coli isolates from multiple vertebrate hosts. A total of 56 multilocus sequence types were identified among the isolates. Five sequence types-ST10, ST38, ST69, ST117, and ST131-were observed among both urine and poultry isolates. Using the Bayesian latent class model, we estimated that 19% (10/52) of the clinical E. coli isolates in our population were foodborne zoonotic strains. These data suggest that a substantial portion of pediatric UTIs in the Washington DC region may be caused by E. coli strains originating in food animals and likely transmitted via contaminated poultry meat.IMPORTANCEEscherichia coli UTIs are a heavy public health burden and can have long-term negative health consequences for pediatric patients. E. coli has an extremely broad host range, including humans, chickens, turkeys, pigs, and cattle. E. coli derived from food animals is a frequent contaminant of retail meat products, but little is known about the risk these strains pose to pediatric populations. Quantifying the proportion of pediatric UTIs caused by food-animal-derived E. coli, characterizing the highest-risk strains, and identifying their primary reservoir species could inform novel intervention strategies to reduce UTI burden in this vulnerable population. Our results suggest that retail poultry meat may be an important vehicle for pediatric exposure to zoonotic E. coli strains capable of causing UTIs. Vaccinating poultry against the highest-risk strains could potentially reduce poultry colonization, poultry meat contamination, and downstream pediatric infections.

Neonatal Amino Acids and Acylcarnitines Associated With Maternal Blood Glucose Levels Throughout Pregnancy: Insights From the Beijing Birth Cohort Study.

To determine the association between maternal blood glucose patterns throughout pregnancy and neonatal amino acids and acylcarnitines. We conducted a prospective cohort study involving 11,457 singleton pregnant women without preexisting diabetes from the Beijing Birth Cohort Study, along with their neonates born between July 2021 and October 2022 in Beijing, China. Distinct maternal glucose trajectories were identified using a latent class model based on blood glucose levels across the three trimesters, and their association with neonatal circulating metabolites, including 11 amino acids and 33 acylcarnitines, was examined, adjusting for potential confounding factors. Three distinct groups of maternal glucose trajectories were identified: consistent normoglycemia (n = 8,648), mid-to-late gestational hyperglycemia (n = 2,540), and early-onset hyperglycemia (n = 269). Mid-to-late gestational hyperglycemia was associated with decreased levels of amino acids (alanine, arginine, ornithine, and proline) involved in the arginine and proline metabolism and urea cycle pathway, as well as increased levels of C4DC+C5-OH and decreased level of C6DC and C10:1. Early-onset hyperglycemia was associated with elevated levels of free acylcarnitine and C4DC+C5-OH and a decreased level of C10:1, involved in the fatty acid oxidation pathway. However, these associations were primarily observed in male neonates rather than in female neonates. Our findings revealed a significant link between maternal glucose trajectories throughout pregnancy and neonatal arginine and proline metabolism, urea cycle pathway, and fatty acid oxidation pathway. These results highlight the importance of maintaining optimal blood glucose levels throughout pregnancy to promote healthy neonatal metabolic outcomes.

Identifying the losers in the transport transition: evidence from Germany

Public acceptance of transport transition policies in the European Union is low as they are considered regressive. This paper provides a clear way of identifying the winners and losers of these policies by focusing on transport poverty. A two-step methodology is followed. First, principal component analysis (PCA) unifies the literature by defining the main underlying dimensions of transport poverty. These highlight the importance of a household’s spatial matching, self-imposed driving restrictions, and available resources. Second, the use of a latent class model (LCM) makes it possible to classify households according to four profiles in the first-ever transport poverty scale (TPS). 14.7 million German households are classified as transport-poor and car-dependent. These two classes represent the most constrained households in terms of resources (time and money) and options available for transport. The degree of spatial matching plays a key role in defining class assignments. Finally, the application of the TPS quantifies the heterogeneous reaction of each transport poverty class to increases in the cost of driving (e.g., the carbon tax). The car-dependent and the transport-poor are the losers in the transition due to inelastic driving demand. Policies reflecting class-based responses to achieve the goals of the transport transition are presented. Alleviating the burden on the transport-poor and car-dependent could increase the acceptability of policies and accelerate the transport transition.

Latent profile analysis of college students’ mental Huagong On the Horizon of Chinese Culture

Objective: To explore the latent class characteristics of college students’ mental Huagong. Methods: A questionnaire survey was carried out on 510 college students by using Collage Mental Huagong Scale(CMHGS) and the Self-assessment Scale of College Students’ Innovation and Entrepreneurship Ability(CSIEASAS). Results: The college students’ mental Huagong can be divided into high Huagong group (40%), high-level low will group (6%), medium-level low emotion group (48%) and low Huagong group (5%). There were significant differences in the scores of entrepreneurial personality, basic entrepreneurial ability, core entrepreneurial ability and social adaptability among the four Latent class model of collage students’ mental Huagong. Conclusion: There are four Latent class model of college students’ mental Huagong , and there are significant differences in each dimensions of college students’ abilities of innovation and entrepreneurship among each latent class models of college students’ mental Huagong.

Generating contingency tables with fixed marginal probabilities and dependence structures described by loglinear models

We present a method to generate contingency tables that follow loglinear models with prescribed marginal probabilities and dependence structures. We make use of (loglinear) Poisson regression, where the dependence structures, described using odds ratios, are implemented using an offset term. Other statistical models related to loglinear models that fall into the scope of this paper, such as the logistic regression model, the latent class model and the extended Rasch are discussed as well. We apply this methodology to carry out simulation studies in the context of population size estimation using dual system and triple system estimators, popular in official statistics. These estimators use contingency tables that summarize the counts of elements enumerated or captured within lists that are linked. The simulation is used to investigate these estimators in the situation that the model assumptions are fulfilled, and the situation that the model assumptions are violated.

Geographic authentication of argentinian teas by combining one-class models and discriminant methods for modeling near infrared spectra

In this study, 110 tea samples from South American countries (Argentina, Brazil, and Paraguay) and Asian countries (India and China) were analyzed using near-infrared spectroscopy (NIRS) together with a two-step chemometric authentication strategy (class modeling techniques and discriminant analysis) to authenticate commercial teas from Argentina. In the first step, one-class models were built and validated to authenticate South American teas using preprocessed NIRS data. For this purpose, data-driven soft independent modeling of class analogy (DD-SIMCA) and one-class partial least squares (OC-PLS) were used. The DD-SIMCA model gave the best results, with a sensitivity of 93.10%, specificity of 100%, and efficiency of 95.00%. In the second step, a support vector machine (SVM) was used to build and validate a multiclass model to discriminate between tea samples from Argentina and neighboring countries of South America. The best model was the combination of nine variables selected by the fast correlation-based filter (FCBF) method, with an accuracy of 98.30%. Therefore, we conclude that the combination of NIRS and two-step chemometric tools can be used to authenticate the geographical origin of samples with high inter-class similarity.

Understanding dairy farmers' trade‐offs between environmental, social and economic sustainability attributes in feeding systems: The role of farmers' identities

AbstractThere is scope for improving the sustainability of intensive dairy farms through the uptake of sustainable production practices such as more grass‐based feeding systems. Such feeding systems can reduce feed‐food competition and the environmental impacts of feed production, among other farm‐level and societal benefits. However, empirical research on how farmers' feed choices mis(align) with sustainability transitions and the associated drivers is limited. This paper explores the trade‐offs that farmers make between the environmental, social and economic sustainability impacts of grass‐based feeding systems based on data from Swedish dairy farmers. Using an identity‐based utility framework and a hybrid latent class model, we find substantial heterogeneity in dairy farmers' trade‐offs between feed‐related sustainability attributes: greenhouse gas emissions, biodiversity, animal welfare, feed self‐sufficiency, feed cost and milk yield. Furthermore, our findings demonstrate that farmers who are strongly interested in the environmental and social sustainability impacts of their dairy feeding systems, beyond economic gains, are motivated mainly by their pro‐environmental and pro‐social identities. Overall, our findings imply that identity‐enhancing interventions are promising policy instruments for encouraging the uptake of more grass‐based feeding systems.

Design and practice of secondary mathematics teaching based on PAD class model

Design and practice of secondary mathematics teaching based on PAD class model

BAYESIAN NESTED LATENT CLASS MODELS FOR CAUSE-OF-DEATH ASSIGNMENT USING VERBAL AUTOPSIES ACROSS MULTIPLE DOMAINS.

Understanding cause-specific mortality rates is crucial for monitoring population health and designing public health interventions. Worldwide, two-thirds of deaths do not have a cause assigned. Verbal autopsy (VA) is a well-established tool to collect information describing deaths outside of hospitals by conducting surveys to caregivers of a deceased person. It is routinely implemented in many low- and middle-income countries. Statistical algorithms to assign cause of death using VAs are typically vulnerable to the distribution shift between the data used to train the model and the target population. This presents a major challenge for analyzing VAs, as labeled data are usually unavailable in the target population. This article proposes a latent class model framework for VA data (LCVA) that jointly models VAs collected over multiple heterogeneous domains, assigns causes of death for out-of-domain observations and estimates cause-specific mortality fractions for a new domain. We introduce a parsimonious representation of the joint distribution of the collected symptoms using nested latent class models and develop a computationally efficient algorithm for posterior inference. We demonstrate that LCVA outperforms existing methods in predictive performance and scalability. Supplementary Material and reproducible analysis codes are available online. The R package LCVA implementing the method is available on GitHub (https://github.com/richardli/LCVA).

Patient Preferences for Treatment in Relapsed/Refractory Acute Leukemia in the United Kingdom: A Discrete Choice Experiment.

Acute leukemia is a cancer of the white blood cells which progresses rapidly and aggressively. There are two types: acute lymphoblastic leukemia (ALL) and acute myeloid leukemia (AML). The latter has a rare subtype: acute promyelocytic leukemia (APL). For some patients, following first-line treatment, remission is not achieved ("refractory disease"), and for others the leukemia returns after achieving remission ("relapse"). For these individuals, outcomes are typically poor. It is, therefore, important to understand patients' treatment priorities in this context. Building upon formative qualitative research, an online survey containing a discrete choice experiment (DCE) was designed to explore patients' treatment preferences in the relapsed/refractory setting. The DCE attributes were mode of administration; quality of life during treatment; chance of response; duration of response; and quality of life during response. Each respondent completed twelve scenarios containing two hypothetical treatments. Participants were eligible if they lived in the United Kingdom and had a diagnosis of acute leukemia. The data were analysed using a latent class model. A total of 95 patients completed the survey. The latent class analysis identified two classes. For both, chance of response was the most important attribute. For class 1, every attribute was important, whereas for class 2, the only important attributes were quality of life (during treatment and response) and chance of response. A greater proportion of respondents would fall into class 1 overall, and those with ALL or APL and those more recently diagnosed were more likely to be in class 2. Our results indicate that patients are strongly concerned about the chance of response, as well as quality of life (to a lesser extent), when faced with different treatment options in the relapsed/refractory setting. However, there is significant preference heterogeneity within the patient population, and other treatment characteristics also matter to many.

Meta-Analisis investigasi model kelas terbalik pada keterampilan berpikir tingkat tinggi (HOTS) siswa matematika: Analisis efek gabungan dan heterogenitas

This research aims to test the effectiveness of the flipped classroom model on students' higher order thinking abilities (HOTS) in mathematics subjects. The research method uses meta-analysis. The sample in this study was 18 primary studies that met the specified inclusion criteria. Data analysis was performed using comprehensive meta-analysis (CMA) software. The results of the analysis using the random-effects approach showed that the combined effect size was (g = 1.82; p < 0.01) (very large effect). This shows that the implementation of the flipped classroom model has a big influence on students' mathematics HOTS when compared to the traditional learning model. The results of the moderator variable analysis show that the effect of implementing the flipped class model on students' mathematics HOTS differs significantly based on the publication type variable (Qb = 7.94; p = 0.01). However, it was not significantly different based on the variables of education level (Qb = 7.45; p = 0.06), experimental group capacity (Qb = 0.29; p = 0.62), and year of publication (Qb = 0.21; p = 0.64). These findings can be the basis for choosing a mathematics learning model so that the quality of mathematics learning using the flipped classroom model can be improved.

The pathways of technology and class models of English language learning: reflecting the new normal era

This research aims to showcase English language learning in the new normal era from 17 May 2022 to 30 June 2023. It focuses on conditions, class models, the reasons for class model adjustment, and the role of technology. Using an open-ended questionnaire, the researchers obtained information about the phenomenon and complete data from 83 teachers in Salatiga and Semarang Regency, Central Java, Indonesia. The study findings demonstrate that the conditions of English language learning in the new normal era coped with learning adaptation, challenging learning, learning improvement, changes in character, the amusement of face-to-face classes, and low understanding of the material. Regarding class models, during two years of the pandemic, teachers predominantly used offline, blended, hybrid, and online class models. The main reasons for the adjustment in the class models were the need for social interaction, knowledge development, behavioral education, and health factors. Regarding technology’s contributions, it was crucial to develop learning materials. This study concludes with empirical insights into consideration for further development of English language learning by integrating technology and various class models so that teachers can mitigate and sustain learning due to probable extenuating circumstances in the future.

Trajectories and predictors of emotional exhaustion in clinical nurses in the context of healthcare crisis: a case study in Xi’an, China

This study investigated the growth trajectory of emotional exhaustion and its predictors in clinical nurses in the context of healthcare crisis in the post pandemic era. A total of 422 Chinese clinical nurses were followed up three times over two months from December 2022 to January 2023, and data from the final 358 clinical nurses were analyzed. The growth mixed model was used to identify the trajectory categories, and the predictive factors of the trajectory types were analyzed by logistic regression. The relationship of the trajectory types of emotional exhaustion with psychological capital were analyzed by variance analysis. The best-fit growth mixture modeling revealed three class models: Class 1 characterized by high and increasing levels of emotional exhaustion, Class 2 characterized by moderate and decreasing levels of emotional exhaustion, and Class 3 characterized by low and decreasing levels of emotional exhaustion. These classes accounted for 9.78%, 83.52%, and 6.70% of clinical nurses, respectively. The results of the univariate analysis indicated that age, years of working experience, gender, past participation in emergency public health event rescue, and sleep quality were associated with the trajectory of emotional exhaustion. Logistic regression analysis revealed that gender, past participation in emergency public health event rescue, and sleep quality were independent predictors of the emotional exhaustion trajectory. The variance analysis of the four dimensions of psychological capital showed that there were statistically significant differences in scores of self-efficacy and hope dimensions among the three groups of emotionally exhausted individuals. Our study demonstrated the heterogeneity of emotional exhaustion among clinical nursing personnel in coping with healthcare system overload. The psychological capital of nursing staff with different categories of emotional exhaustion varies. Managers should actively pay attention to the predictive role of gender, past participation in emergency public health event rescue, and sleep quality on the development of emotional exhaustion in clinical nurses.

Adaptive dual teacher incremental learning for defect detection of flexible printed circuit

Flexible Printed Circuit (FPC) is a high space utilization and high flexibility circuit board. Due to the flexible production method of FPC, new class defects will constantly appear during the production process. Existing methods for incremental class learning can enable the model to learn new class defects, however less attention is paid to the balance between old and new class defects. In addition, the knowledge of new class defects extracted by old class teacher is inaccurate because the teacher model has not learned new class defects. To address these issues, this paper proposes a model called Adaptive Dual Teacher Incremental Learning with Decoupled Feature Distillation for defect detection (ADT-ILDFD), which employs an adaptive dual-teacher model to adjusts the importance of old and new classes through adaptive modular learning during the training process, and thus better accommodate both old and new classes simultaneously in different incremental environments. In addition, decoupled feature distillation was designed, which decouples the feature knowledge in order to distill the feature knowledge extracted by old class model more accurately. Finally, we constructed the FPC defect dataset named FPCSD2023 and conducted experiments on both FPCSD2023 and NEU-DET to demonstrate the effectiveness of ADT-ILDFD.

Training Vehicle Detection and Classification Models with Less Data: An Active Learning Approach

The application of computer vision in transportation engineering has facilitated real-time traffic flow optimization, vehicle counting, anomaly detection, and ameliorated transportation safety. Most vision systems are, however, developed through a supervised learning process, which can be data hungry and costly because it requires manual annotation of objects from a variety of sources. The general rule of thumb for building accurate and transferrable vision models has been to increase the quality, diversity, and quantity of the annotated datasets used in model training. This paper presents a simple, yet efficient active learning framework that significantly reduces the number of annotations needed to build a state-of-the-art vehicle detection and classification model. To achieve this, we first leverage a vision transformer that generates embeddings rich with information needed to quantify the similarity and diversity between images in a two-dimensional embedding space. To select which images from the embedding space should be annotated, we propose a scoring and sampling strategy that minimizes class imbalance and model uncertainty through an iterative process. The latest iteration of the You Only Look Once (YOLO) model, YOLOv8, is used as the active learner. We compare the efficacy of our proposed active learning methods with models developed at much higher sampling rates using the mean average precision. The models developed were also integrated with tracking algorithms to evaluate differences in accuracy for vehicle counts and their practical implications for direction counts.

Research on Optimization of International Multimodal Transportation Routes Based on Uncertainty Theory

Delays in international multimodal marine transport can result in subsequent connection delays, leading to prolonged transportation time and increased uncertainty. To identify more reliable and cost-effective robust transportation routes, this study establishes a constrained planning model for multimodal transportation routes. Employing uncertainty theory, the model is transformed into a deterministic equivalent class model and ultimately incorporates an adaptive differential evolution (ADE) algorithm. Results of a case study indicate that: (i) the proposed model, compared with deterministic models, exhibits greater robustness and better aligns with practical transportation scenarios, resulting in substantial reductions in actual transportation time and cost; (ii) the solution efficiency of the ADE algorithm surpasses that of the genetic algorithm and Dijkstra algorithm; (iii) the start time of transportation and the confidence level of uncertainty also play crucial roles in influencing route selection. Therefore, decision-makers should consider a multifaceted approach when formulating transportation routes.

Consumers’ preferences for the attributes of plant-based meat in China: A best-worst scaling approach

Plant-based meat could contribute substantially to nutritional security and environmental sustainability. Nevertheless, its diffusion is highly dependent on consumers’ perception and acceptance. Plant-based meat products have various attributes through which nutritional, environmental, and ingredient information are delivered to consumers. To elucidate consumers’ preferences for these attributes, we conducted a large-scale consumer survey in five major cities of the largest national meat consumption market, China, with a total of about 2500 respondents. Best-worst count analysis indicated that “zero cholesterol” is the most preferred and “vegan formula” is the least preferred attribute. Mixed logit model analysis showed that “rich in dietary fiber”, “zero hormones”, and “zero cholesterol” are the three most preferred attributes. Latent class model analysis revealed that respondents with better socioeconomic conditions tended to focus more on the positive nutritional aspects of plant-based meat, whereas respondents with worse conditions tended to place a higher value on avoiding negative health impacts of traditional meat when choosing plant-based meat. This study suggests that Chinese consumers pay more attention to nutritional attributes as compared to other attributes. In addition, apparent market segments exist among different consumers. The findings of this study could contribute to the transformation of sustainable food system.

Islam and the Transmission of Cultural Identity in Four European Countries

Abstract Studies exploring the integration of European immigrants tend to find cultural gaps between Muslim children and their peers. While some scholars argue that parent-to-child transmission is a key mechanism underlying this pattern, others privilege extrafamilial explanations by pointing to differences in cultural values within Muslim households. In the present study, I argue that these mixed results stem from a tendency in the literature to analyze distinct components of personal culture in isolation from cognate dimensions. To address this shortcoming, I use multigroup latent class models to explore how a wide range of attitudes (tapping ethnocultural identity, gender norms, sexual liberalism, and perspectives on integration) are clustered together in disparate regions of the belief space, marking distinct cultural identities. Then, I fit a series of logistic regressions to map how these cultural identities are distributed among immigrant-origin samples in four European countries and transmitted across generational lines. Ultimately, I arrive at the following conclusion: while Muslim youth stand out from their peers vis-à-vis their cultural identity profiles, there is little evidence to suggest that this pattern is decisively shaped by parent-to-child transmission.

Exploring preferences for contractual terms in a scenario of ecological transition for the agri-food sector: a latent class approach

Governance mechanisms along the agri-food supply chains are increasingly important in a scenario of ecological transition. Under the conceptual and analytical lens of the Transaction Cost Economics, we explored farmers’ preferences towards a variety of clauses usually adopted in production contracts. To this purpose, a discrete choice experiment among 190 durum wheat producers in Italy was conducted. Results from a latent class model showed that producers were mainly interested in fixed prices formula and to join shared rules of production but revealed little or no interest for compelling sustainable cultivation techniques and the provision of technical assistance. However, these preferences are heterogeneous across farmers and vary depending on their level of education and previous use of contractual arrangements, with relevant implications for contract design and management. Governance mechanisms along the agri-food supply chains are increasingly important in a scenario of ecological transition. Under the conceptual and analytical lens of the Neo Institutional Economics, we explored farmers’ preferences towards a variety of clauses usually adopted in production contracts. To this purpose, a discrete choice experiment among 190 durum wheat producers in Italy was conducted. Results from a latent class model showed that that producers were mainly interested in fixed prices formula and to join shared rules of production but revealed little or no interest for compelling sustainable cultivation techniques and the provision of technical assistance. However, these preferences are heterogenous across farmers and vary depending on their level of education and previous use of contractual arrangements, with relevant implications for contract design and management.