Search

BACKGROUND: Since the inception of the Youth Risk Behavior Surveillance System in 1991, all surveys have been conducted in schools, using paper and pencil instruments (PAPI). For the 2019 YRBSS, sites were offered the opportunity to conduct their surveys using electronic data collection. This study aimed to determine whether differences in select metrics existed between students who completed the survey electronically versus using PAPI. METHODS: Thirty risk behaviors were examined in this study. Data completeness, response rates and bivariate comparisons of risk behavior prevalence between administration modes were examined. RESULTS: Twenty‐nine of 30 questions examined had more complete responses among students using electronic surveys. Small differences were found for student and school response rates between modes. Twenty‐five of 30 adolescent risk behaviors showed no mode effect. CONCLUSIONS: Seven of 44 states and DC participated electronically. Because survey data were more complete; school and student response rates were consistent; and minor differences existed in risk behaviors between modes, the acceptability of collecting data electronically was demonstrated. [ABSTRACT FROM AUTHOR]

In unsupervised domain adaptation (UDA), knowledge is transferred from label-rich source domains to relevant but unlabeled target domains. Current most popular state-of-the-art works suggest that performing domain alignment from the class perspective can alleviate domain shift. However, most of them based on domain adversarial which is hard to train and converge. In this paper, we propose a novel contrastive learning to improve diversity and discriminability for domain adaptation, dubbed as IDD_ICL, which improve the discriminativeness of the model while increasing the sample diversity. To be precise, we first design a novel implicits contrastive learning loss at sample-level by implicit augment sample of the source domain. While augmenting the diversity of the source domain, we can cluster the samples of the same category in the source domain together, and disperse the samples of different categories, thereby improving the discriminative ability of the model. Furthermore, we show that our algorithm is effective by implicitly learning an infinite number of similar samples. Our results demonstrate that our method doesn't require complex technologies or specialized equipment, making it readily adoptable and applicable in practical scenarios. [ABSTRACT FROM AUTHOR]

The fast warning for financial risk of enterprises has always been a realistic demand for their managers. Currently, this mainly relies on expert experience to make comprehensive analysis from massive business data. Benefitting from the strong computational performance of deep learning, this paper proposes a fuzzy neural network (FNN)-based intelligent warning method for financial risk of enterprises. An improved FNN structure with time-varying coefficients and time-varying time lags is established to extract features of enterprises from complex financial context. The algorithm of fuzzy C-means and fuzzy clustering based on sample data are studied. In this paper, the fuzzy C-means algorithm is used to cluster the samples, the input sample set is preprocessed, a new set of learning samples is formed, and then the neural network is trained. The enterprise financial risk sample and its modular FNN model are established, and the evaluation of the enterprise financial risk sample is simulated. Then, a decision part is added following the FNN part to output the warning results. After that, we have also conducted a case study as simulation experiments to evaluate the proposed technical framework. The obtained results show that it can perform well in the fast warning of financial risk for enterprises. [ABSTRACT FROM AUTHOR]

Unsupervised hashing for cross-modal retrieval has received much attention in the data mining area. Recent methods rely on image-text paired data to conduct unsupervised cross-modal hashing in batch samples. There are two main limitations for existing models: (1) learning of cross-modal representations is restricted to batches; (2) semantically similar samples may be wrongly treated as negative. In this paper, we propose a novel category-level contrastive learning for unsupervised cross-modal hashing, which alleviates the above problems and improves cross-modal query accuracy. To break the limitation of learning in small batches, a selected memory module is first proposed to take global relations into account. Then, we obtain pseudo labels through clustering and combine the labels with the Hadamard Matrix for category-centered learning. To reduce wrong negatives, we further propose a memory bank to store clusters of samples and construct negatives by selecting samples from different categories for contrastive learning. Extensive experiments show the significant superiority of our approach over the state-of-the-art models on MIRFLICKR-25K and NUS-WIDE datasets. [ABSTRACT FROM AUTHOR]

Sample clustering techniques play a crucial role in the data‐driven state evaluation of electromechanical equipment, and selecting an appropriate similarity measurement method for sample sets helps improve the clustering performance. The Jaccard coefficient is a commonly employed indicator of similarity for scalar set‐type samples. In this paper, we propose an incremental clustering algorithm for matrix‐type samples by defining an improved Jaccard coefficient. First, a new binary relation is formulated to derive a relationship matrix between samples. Second, an undirected graph is given by using the relationship matrix, and an improved pruning operation is provided to simplify the graph by eliminating redundant edges. Then, a new relationship matrix is generated according to the modified graph, which enables the calculation of the improved Jaccard coefficient. By using the improved Jaccard coefficient, the improved incremental clustering algorithm updates cluster centers by selecting a particular sample to maximize the sum of similarities between the selected sample and other samples within the same cluster. Finally, the effectiveness of the proposed incremental clustering algorithm is demonstrated in fault diagnosis and remaining useful life estimation application scenarios, respectively. The experimental results indicate that the improved algorithm outperforms traditional clustering methods. [ABSTRACT FROM AUTHOR]

In this paper, we propose a method for the generation of simulated test scenarios for autonomous driving. Based on the requirements of standard regulatory test scenarios, we can generate virtually simulated scenarios and functional scenario libraries for autonomous driving, which can be used for the simulated verification of different ADAS functions. Firstly, the operational design domain (ODD) of a functional scenario is selected, and the weight values of the ODD elements are calculated. Then, a combination test algorithm based on parameter weights is improved to generate virtually simulated autonomous driving test cases for the ODD elements, which can effectively reduce the number of generated test cases compared with the traditional combination test algorithm. Then, the traffic participant elements in each test case are sampled and clustered so as to obtain hazard-specific scenarios. Then, the values of the subelements under the traffic participant element in each test case are sampled and clustered to obtain hazard-specific scenarios. Finally, the specific scenarios are applied to the automatic emergency braking (AEB) system on the model-in-the-loop (MIL) testbed to verify the effectiveness of this scenario generation method. [ABSTRACT FROM AUTHOR]

Background: Indigenous Maya women in the rural highlands of Guatemala have traditionally faced constraints to decision-making and participation in community affairs. Anecdotal experiences from previous Curamericas Global projects in Guatemala and Liberia have suggested that interventions using the CBIO+ Approach (which consists of implementing together the Census-Based, Impact-Oriented Approach, the Care Group Approach, and Community Birthing Centers), can be empowering and can facilitate improvements in maternal and child health. This paper, the eighth in a series of 10 papers examining the effectiveness of CBIO+ in improving the health and well-being of mothers and children in an isolated mountainous rural area of the Department of Huehuetenango, explores changes in women's empowerment among mothers of young children associated with the Curamericas/Guatemala Maternal and Child Health Project, 2011–2015. Methods: Knowledge, practice, and coverage (KPC) surveys and focus group discussions (FGDs) were used to explore six indicators of women's empowerment focusing on participation in health-related decision-making and participation in community meetings. KPC surveys were conducted at baseline (January 2012) and endline (June 2015) using standard stratified cluster sampling. Seventeen FGDs (9 with women, 3 with men, 2 with mothers-in-law, and 3 with health committees), approximately 120 people in all, were conducted to obtain opinions about changes in empowerment and to identify and assess qualitative factors that facilitate and/or impede women's empowerment. Results: The KPC surveys revealed statistically significant increases in women's active participation in community meetings. Women also reported statistically significant increases in rates of participation in health-related decision-making. Further, the findings show a dose-response effect for two of the six empowerment indicators. The qualitative findings from FGDs show that the Project accelerated progress in increasing women's empowerment though women still face major barriers in accessing needed health care services for themselves and their children. Conclusion: The Project achieved some notable improvements in women's decision-making autonomy and participation in community activities. These improvements often translated into making decisions to practice recommended health behaviors. Traditional cultural norms and the barriers to accessing needed health services are not easily overcome, even when empowerment strategies are effective. [ABSTRACT FROM AUTHOR]

Background: Community health worker programmes have the potential to contribute critically towards universal health coverage. However, CHWs globally have often continued to operate on the periphery of the health care system, viewed as a non-essential cadre. This results in a workforce that often remains disempowered and under-supported. This paper presents evidence from a study conducted in a rural part of South Africa, to better understand issues of CHW prioritisation, integration, and empowerment. Methods: We applied an analytical lens based on empowerment theory and conducted a secondary analysis of qualitative data emerging from a sub-study of a cRCT evaluating the effectiveness of supportive supervision for CHWs within a large-scale national CHW programme. The cRCT was conducted between 2017 and 2022, and 39 CHWs were included in the study. Results: We organised our findings across the four domains of structural empowerment; information, resources, support, and opportunity, and mapped these domains against the domains of psychological empowerment. Our findings show how CHWs are still working in the periphery of the healthcare system. Without sufficient prioritisation, high level-support from national and district governments, and sufficient investments in programmatic domains—such as training, equipment, and supportive supervision—it is likely that the CHW cadre will continue to be seen as informal health care workers. Conclusions: CHW empowerment could be a lever to potentially transform the current health system towards universal coverage; however, this process can only happen with sufficient high-level prioritization and investment. [ABSTRACT FROM AUTHOR]

Selecting emergency suppliers from a wide range of candidates based on their performance under each criterion can be regarded as a multi-criterion decision making (MCDM) problem. Existing MCDM models to solve the emergency supplier selection problem ignored situations where large-scale suppliers exist, the influence of criteria weights on the robustness of ranking results, and the influence of psychology of regret aversion and disappointment aversion on decision results. To make up for these deficiencies, this paper proposes an MCDM model to solve emergency supplier selection problem with large-scale alternatives. Firstly, to avoid the influence of criteria weights on ranking of alternatives, the Robustness, Correlation, and Standard Deviation (ROCOSD) method is introduced to determine objective weights of criteria based on three objectives. Secondly, the τ-balanced clustering method is applied to cluster large-scale alternatives into balanced clusters. Next, considering the psychology of regret aversion and disappointment aversion of decision makers, a two-stage method is proposed to rank alternatives, which identifies the optimal alternative within each cluster and forms a new cluster consisting of these optimal alternatives in the first stage, and selects the optimal alternative from the new-formed cluster in the second stage. A numerical case is given to validate the proposed model. [ABSTRACT FROM AUTHOR]

Federated learning has gained significant attention due to its groundbreaking ability to enable distributed learning while maintaining privacy constraints. However, as a consequence of data heterogeneity among decentralized devices, it inherently experiences significant learning degradation and slow convergence speed. Therefore, it is natural to employ the concept of clustering homogeneous clients into the same group, allowing only the model weights within each group to be aggregated. While most existing clustered federated learning methods employ either model gradients or inference outputs as metrics for client partitioning to group similar devices together, heterogeneity may still exist within each cluster. Moreover, there is a scarcity of research exploring the underlying reasons for determining the appropriate timing for clustering, resulting in the common practice of assigning each client to its own individual cluster, particularly in the context of highly non-independent and identically distributed (Non-IID) data. In this paper, we introduce a two-stage decoupling federated learning algorithm with adaptive personalization layers named FedTSDP, where client clustering is performed twice according to inference outputs and model weights, respectively. Hopkins amended sampling is adopted to determine the appropriate timing for clustering and the sampling weight of public unlabeled data. In addition, a simple yet effective approach is developed to adaptively adjust the personalization layers based on varying degrees of data skew. Experimental results show that our proposed method has reliable performance on both IID and non-IID scenarios. [ABSTRACT FROM AUTHOR]

This paper proposes two classes of robust ratio type estimators of finite population mean and two classes of robust ratio type estimators of finite population variance using a single auxiliary variable under the adaptive cluster sampling design. Seven robust ratio type estimators have been developed from each class. The generalized expressions of bias and mean square error for each class have been derived up to the first order of approximation. The exact expressions of bias and MSE for all the developed estimators have been presented. Using simulation studies conducted in R programming, the high efficiency of all the developed estimators over similar existing estimators in adaptive cluster sampling has been demonstrated. [ABSTRACT FROM AUTHOR]

Purpose: The paper identifies the influence of socio-economic factors and livelihood training in stimulating micro-entrepreneurship among women self-help group (SHG) members. Design/methodology/approach: The study is based on a sample of 416 women SHG members drawn from all the four districts of Sikkim using cluster sampling procedure. A multivariate binary logistic model is used to find the impact of socio-economic factors, and a Poisson regression has been used to find the impact of training on fostering micro-entrepreneurship. The result is validated using a propensity score matching approach which corrects for the potential self-selection bias in the sample. Subsequently, a covariate adjustment estimator verifies the robustness of the approach. Findings: The study finds that "size of landownership", "amount of loan borrowed", "member's age", "number of earning and dependent members", "number of years of SHG enrolment" as well as the "district to which the member belongs to" have a statistically significant influence on the graduation of SHG members to micro-entrepreneurs. Furthermore, it is found that members attending the livelihood training programmes had a significantly higher number of microenterprises. Originality/value: The study differentiates itself by providing empirical evidence on how socio-economic factors and livelihood training stimulate micro-entrepreneurship among SHG women of Sikkim, which has so far remained unexplored. Moreover, advanced econometric method has been used to eliminate the possible self-selection bias involved with training participation and thereby provides reliable and robust results. Peer review: The peer review history for this article is available at: https://publons.com/publon/10.1108/IJSE-01-2023-0070 [ABSTRACT FROM AUTHOR]

The complexity of the data type and distribution leads to the increase in uncertainty in the relationship between samples, which brings challenges to effectively mining the potential cluster structure of data. Ensemble clustering aims to obtain a unified cluster division by fusing multiple different base clustering results. This paper proposes a three-way ensemble clustering algorithm based on sample's perturbation theory to solve the problem of inaccurate decision making caused by inaccurate information or insufficient data. The algorithm first combines the natural nearest neighbor algorithm to generate two sets of perturbed data sets, randomly extracts the feature subsets of the samples, and uses the traditional clustering algorithm to obtain different base clusters. The sample's stability is obtained by using the co-association matrix and determinacy function, and then the samples can be divided into a stable region and unstable region according to a threshold for the sample's stability. The stable region consists of high-stability samples and is divided into the core region of each cluster using the K-means algorithm. The unstable region consists of low-stability samples and is assigned to the fringe regions of each cluster. Therefore, a three-way clustering result is formed. The experimental results show that the proposed algorithm in this paper can obtain better clustering results compared with other clustering ensemble algorithms on the UCI Machine Learning Repository data set, and can effectively reveal the clustering structure. [ABSTRACT FROM AUTHOR]

The health of young people is crucial for the future and development of a nation. It is the collective responsibility and imperative mission of society to ensure the holistic well-being, both physically and mentally, of young individuals. Therefore, it is essential to thoroughly comprehend the factors that influence their health in order to expedite the exploration of effective solutions. The objective of this study is to comprehend the mechanisms that underlie the correlation between physical exercise behavior and psychological resilience among teenagers, while also examining the mediating role played by social sensitivity and need to belong. So put forward the hypothesis: (1) physical exercise behavior can positively predict the psychological resilience. (2) Social sensitivity and need to belong plays a mediating role between physical exercise behavior and psychological resilience. (3) Social sensitivity and need to belong plays a chain mediating role between physical exercise behavior and psychological resilience. Using the cluster sampling method, a total of 1106 students (with an average age of 15.7 and a standard deviation of 0.598) who met the requirements were surveyed from Shandong Province in China. Standard scales were utilized to assess Physical Exercise Behavior, Psychological Resilience, Social Sensitivity, and Need to Belong. For data analysis, Pearson's correlation analysis and bias-corrected percentile Bootstrap method were sequentially conducted. (1) The present study did not find any significant methodological bias, and the observed correlations between physical exercise behavior, psychological resilience, social sensitivity, and need to belong were all statistically significant; (2) Based on the self-determination theory, this study elucidates the relationship between physical exercise behavior and psychological resilience among teenagers. The findings indicate that physical exercise behavior positively predicts the need to belong and psychological resilience, while negatively predicting social sensitivity. Similarly, social sensitivity negatively predicts the need to belong and psychological resilience. Moreover, the need to belong directly and positively predicts psychological resilience. Importantly, all hypotheses proposed in this paper were empirically supported. (3) The indirect effect of the path mediated by social sensitivity is 0.009, while the indirect effect of the path mediated by need to belong is 0.033. Additionally, the combined indirect effect of both social sensitivity and need to belong as mediating variables is 0.014. (4) The cumulative sum of all these indirect effects amounts to 0.056. Based on the self-determination theory, we propose a chain mediation model, specially, physical exercise behavior can significantly positively predict psychological resilience, among which, social sensitivity and need to belong play a significant mediating role between Physical exercise behavior and psychological resilience. In addition, the adoption of good physical exercise behavior can enhance the psychological resilience of adolescents by diminishing social sensitivity and augmenting the need to belong. [ABSTRACT FROM AUTHOR]

Unsupervised person re-identification (re-ID) aims to identify the same person in unsupervised settings, which is a realistic and challenging problem due to getting rid of the dependence on annotations. Existing pseudo-label-based methods have been proved to be effective by discriminative representation learning and high-quality pseudo labels. However, there are some undesirable similarity structures in images affected by different states (e.g., backgrounds, occlusions, or other target confusions) which may generate noisy labels and lead to model degradation. To tackle the above issues, this paper analyzes the key factors for unsupervised re-ID performance from two perspectives: network structure and loss function. Specifically, we propose a multi-confidence contrastive learning with feature refinement (MCFR) model which can enhance the feature discriminative ability and meanwhile alleviate the impact of noisy data. For network architecture, a feature refined block is embedded into a multi-branch network, constructing FR-net. For loss function, a multi-confidence contrastive loss (MCCL) based on similarity and confidence relationship is developed. MCCL can flexibly update the list of hard negative samples in clusters and outliers, promoting the model by learning from confident samples. Extensive experiments including visualizations and ablation studies validate the effectiveness of each component in the MCFR model. Additionally, compared with the state-of-the-art unsupervised re-ID methods, our method achieves considerable performance on three benchmark datasets. [ABSTRACT FROM AUTHOR]

Direct observation, through surveys, underpins nearly all aspects of environmental sciences. Survey design theory has evolved to make sure that sampling is as efficient as possible whilst remaining robust and fit-for-purpose. However, these methods frequently focus on theoretical aspects and often increase the logistical difficulty of performing the survey. Usually, the survey design process will place individual sampling locations one-by-one throughout the sampling area (e.g. random sampling). A consequence of these approaches is that there is usually a large cost in travel time between locations. This can be a huge problem for surveys that are large in spatial scale or are in inhospitable environments where travel is difficult and/or costly. Our solution is to constrain the sampling process so that the sample consists of spatially clustered observations, with all sites within a cluster lying within a predefined distance. The spatial clustering is achieved by a two-stage sampling process: first cluster centres are sampled and then sites within clusters are sampled. A novelty of our approach is that these clusters are allowed to overlap and we present the necessary calculations required to adjust the specified inclusion probabilities so that they are respected in the clustered sample. The process is illustrated with a real and on-going large-scale ecological survey. We also present simulation results to assess the methods performance. Spatially clustered survey design provides a formal statistical framework for grouping sample sites in space whilst maintaining multiple levels of spatial-balance. These designs reduce the logistical burden placed on field workers by decreasing total travel time and logistical overheads.Supplementary materials accompanying this paper appear on-line. [ABSTRACT FROM AUTHOR]

This paper aims to estimate the prevalence of client-perpetrated violence against female sex workers (FSWs) in Bogotá and to understand what structural and environmental factors are associated with such victimization. The project used secondary data from interviews with 2,684 FSWs. Multivariable binary logistic regression was used to test for associations with client-perpetrated violence. Findings reveal that factors such as experiencing police harassment and social stigma were positively associated with client-perpetrated violence. Situational factors such as providing services in motels, hotels, on the street, and in cars were also associated with increased odds of becoming a victim of violence. [ABSTRACT FROM AUTHOR]

Aim: This study aimed to explore and confirm the factor structure and item distribution of the Chinese version of Self‐care Ability Scale for the Elderly among older Chinese people. Methods: A cluster random sampling method was used to collect 2856 older people from five different cities of China from July 2018 to July 2019. A questionnaire comprised of socio‐demographic information and the Chinese version of the Self‐care Ability Scale for the Elderly was administered, and SPSS 21.0 and Mplus 7.4 were used for analyses. Results: Participants were aged 60 to 92 years, with a mean age of 71.52 (SD = 7.68) years. In the Exploratory Factor Analysis, the three‐factor structure and item distribution of the Chinese version of the Self‐care Ability Scale for the Elderly were verified to be better than other options. The factor loadings varied from 0.428 to 0.800, and the communality values ranged from 0.426 to 0.792. The modified model showed a good fit in the Confirmatory Factor Analysis. Conclusion: The Chinese version of the Self‐care Ability Scale for the Elderly has a clear three‐factor structure and a good item distribution. It serves as a convenient and accurate assessment tool to assess the self‐care ability of older Chinese people. Summary statement: What is already known about this topic? Assessing self‐care ability plays a significant role in the care of older people.The Chinese version of the Self‐care Ability Scale for the Elderly is often used to assess the self‐care ability of older people in China but whilst it has good reliability and content validity, its factor structure and item distribution are not clearly explained in recent literature.Solid evidence on the factor analysis of the Chinese version of the Self‐care Ability Scale for the Elderly is needed to maintain a high standard of health care assessment for the older Chinese people. What this paper adds? We confirmed that the Chinese version of the Self‐care Ability Scale for the Elderly has a clear three‐factor structure and item distribution.The Chinese version of the Self‐care Ability Scale for the Elderly can now be used as the study confirmed its validity and reliability. The implications of this paper: The instrument can be effective in older Chinese people to identify their self‐care support needs and thereby to enable provision of opportunities to implement appropriate nursing care.Health care professionals can use this instrument to assess the alterations in self‐care ability of older Chinese people after particular interventions. [ABSTRACT FROM AUTHOR]

Background: This paper presents the baseline characteristics and their moderators in the Healthy Learning Mind (HLM)– school-based cluster-randomized controlled trial. Objectives: The paper evaluates the state of various measures of well-being, their moderators and how these results compare to national and global norms/population studies. Methods: Data were collected from all participants prior to the intervention and further analyzed by gender, grade and perceived socioeconomic status, including standardized measures for resilience, depressive symptoms and socioemotional functioning; health-related quality of life, dispositional mindfulness, satisfaction with life, compassion/self-kindness, self-rated health and morning tiredness. Results: Participating 2793 students (1425 girls, 1368 boys), ages 12 to 15 years, filled in the questionnaires. The outcomes were in line with previous research, demonstrating gender differentiation and lower wellbeing among older children and adolescents. Conclusions: All outcomes were associated with perceived socioeconomic status, suggesting that perceived low socioeconomic status should be addressed as a serious risk factor and included as a moderator in similar trials. [ABSTRACT FROM AUTHOR]

In this paper, we suggest using a partial randomized response model using Poisson distribution to efficiently estimate a rare sensitive attribute by applying the probability proportional to size (PPS) sampling method when the population is composed of several different and sensitive clusters. We have obtained estimators for a rare and sensitive attribute and their variances and variance estimates by applying PPS sampling and two-stage equal probability sampling. We compare the efficiency between the estimators of the rare sensitive attribute, one obtained via PPS sampling with replacement and the other obtained using the two-stage equal probability sampling with replacement. As a result, it is confirmed that the estimate obtained via the PPS sampling with replacement is more efficient than the estimate provided by the two-stage equal probability sampling with replacement when the cluster sizes are different. [ABSTRACT FROM AUTHOR]

Objective: This paper aims to report South East Asian Nutrition Surveys (SEANUTS) II Malaysia data on nutritional status, dietary intake and nutritional biomarkers of children aged 6 months to 12 years. Design: Cross-sectional survey conducted in 2019–2020. Setting: Multistage cluster sampling conducted in Central, Northern, Southern and East Coast regions of Peninsular Malaysia. Participants: 2989 children aged 0·5–12·9 years. Results: Prevalences of stunting, thinness, overweight and obesity among children aged 0·5–12·9 years were 8·9 %, 6·7 %, 9·2 % and 8·8 %, respectively. Among children below 5 years old, 11·4 % were underweight, 13·8 % had stunting and 6·2 % had wasting. Data on nutritional biomarkers showed that a small proportion of children aged 4–12 years had Fe (2·9 %) and vitamin A deficiencies (3·1 %). Prevalence of anaemia was distinctly different between children below 4 years old (40·3 %) and those aged 4 years and above (3·0 %). One-fourth of children (25·1 %) had vitamin D insufficiency, which was twice as prevalent in girls (35·2 % v. boys: 15·6 %). The majority of children did not meet the recommended dietary intake for Ca (79·4 %) and vitamin D (94·8 %). Conclusions: Data from SEANUTS II Malaysia confirmed that triple burden of malnutrition coexist among children in Peninsular Malaysia, with higher prevalence of overnutrition than undernutrition. Anaemia is highly prevalent among children below 4 years old, while vitamin D insufficiency is more prevalent among girls. Low intakes of dietary Ca and vitamin D are also of concern. These findings provide policymakers with useful and evidence-based data to formulate strategies that address the nutritional issues of Malaysian children. [ABSTRACT FROM AUTHOR]

The purpose of this study was to collect preliminary data for the design of real-time and integrated performance assessments in science practicum at two public universities and two private universities in Eastern Indonesia. The case study design and quantitative and qualitative approaches were adopted in this study. Cluster random sampling technique was used to select four universities, 376 participants respectively from university students and lecturers, four vice deans, four heads of study programs, four heads of laboratories, and four teaching assistants. Data were gathered through questionnaires, in-depth interviews, and observations. The chi-square test was used to analyze the data descriptively. Findings indicated that a performance assessment was not carried out since standard measurement instruments that provided real-time and integrated feedback had not been developed. There was no significant difference (p > 0.05) between public and private universities on the science practicum assessment, implying that both public and private universities still used paper-based practicum assessments. There was also no significant difference between the responses of university students and lecturers on practicum performance assessments (p > 0.05). University students' experience in assessing science practicums was similar to those of lecturers', which suggests that practicum of performance assessments was not real-time and integrated with feedback. This study recommends developing a real-time and integrated performance assessment with automated feedback on the science practicum, in order to provide continuous improvement of students and monitoring their learning progress. [ABSTRACT FROM AUTHOR]

This paper examines the dynamics of learning and how they relate to the optimal teaching strategies for learners. The authors aim to enhance learners' information processing abilities by highlighting the importance of considering different learning theories in the classroom. These considerations are based on the qualitative behaviour of learners' conscious awareness, cognitive and metacognitive engagement. A multi-stage cluster sampling procedure was employed to select a sample of 840 Grade 11 learners from 20 schools in the education district. A questionnaire was used to collect data, and statistical tests were conducted to validate and analyse the data. The study's results suggest that it is crucial for teachers to use a variety of learning theories and to consider variables that facilitate effective information processing by learners. Understanding diverse learning theories and their interplay positively influence the cognitive development of learners. The study's findings, which include SEM and HLM analyses and implications from teachers' application of juxtaposed theories, suggest that inefficiencies may exist in certain teaching approaches and in individual learners' information processing. Teachers who understand and apply the simultaneous impact of juxtaposed theories of learning can enable learners to assimilate or accommodate incoming information into their cognitive schema, leading to deeper and more meaningful understanding. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Economics & Administrative Sciences is the property of Republic of Iraq Ministry of Higher Education & Scientific Research (MOHESR) and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Conversational recommender systems elicit user preference via conversational interactions. By introducing conversational key-terms, existing conversational recommenders can effectively reduce the need for extensive exploration required by a traditional interactive recommender. However, there are still limitations of existing conversational recommender approaches eliciting user preference via key-terms. First, the key-term data of the items needs to be carefully labeled, which requires a lot of human efforts. Second, the number of the human labeled key-terms is limited and the granularity of the key-terms is fixed, while the elicited user preference is usually from coarse-grained to fine-grained during the conversations. In this paper, we propose a clustering of conversational bandits algorithm. To avoid the human labeling efforts and automatically learn the key-terms with the proper granularity, we online cluster the items and generate meaningful key-terms for the items during the conversational interactions. Our algorithm is general and can also be used in the user clustering when the feedback from multiple users is available, which further leads to more accurate learning and generations of conversational key-terms. Moreover, to learn the user clustering structure more efficiently in more complex user clustering structure, we further propose a simple yet effective soft user clustering module to perform exploration on user clustering via sampling the posterior user representations. We analyze the regret bound of our learning algorithm. In the empirical evaluations, without using any human labeled key-terms, our algorithm effectively generates meaningful coarse-to-fine grained key-terms and performs as well as or better than the state-of-the-art baseline. [ABSTRACT FROM AUTHOR]

In this paper, the authors formulated the problem of finding optimum values of cluster size and sampling units for multivariate cluster sampling surveys. In multivariate case, the use of individual optimum allocations is not advisable. For such cases, any suitable criterion is required to decide the acceptable cluster size and sampling unit which may be optimum for all characteristics under study. Moreover, the use of Gamma cost function is introduced for the study of multivariate cluster sampling surveys. The problem is formulated as an All Integer Nonlinear Programing Problem (AINLPP). In this paper, the solution procedures are suggested for solving the formulated AINLPP. The numerical examples show the practical applications of the solution procedures for multivariate cluster sampling design. [ABSTRACT FROM AUTHOR]

Imbalanced data poses a significant challenge in machine learning, as conventional classification algorithms often prioritize majority class samples, while accurately classifying minority class samples is more crucial. The synthetic minority oversampling technique (SMOTE) represents one of the most renowned methods for handling imbalanced data. However, both SMOTE and its variants have limitations due to their insufficient consideration of data distribution, leading to the generation of incorrect and unnecessary samples. This paper, therefore, introduces a novel oversampling algorithm called data distribution and spectral clustering-based SMOTE (DDSC-SMOTE). This algorithm addresses the shortcomings of SMOTE by introducing three innovative data distribution-based improvement strategies: adaptive allocation of synthetic sample quantities strategy, seed sample adaptive selection strategy, and synthetic sample improvement strategy. First, we use the k-nearest neighbor sample labels and the local outlier factor algorithm to remove noisy and outlier samples. Next, we leverage spectral clustering to identify clusters within the minority class and propose a dual-weight factor that considers inter-cluster and intra-cluster distances to allocate the number of synthetic samples effectively, addressing interclass and intraclass imbalances. Furthermore, we introduce a relative position weight coefficient to determine the probability of selecting seed samples within the subcluster, ensuring that important minority samples have higher chances of being sampled. Finally, we improve the SMOTE sample synthesis formula for safer generation. Extensive comparisons on real datasets from the UCI repository demonstrate that DDSC-SMOTE outperforms seven state-of-the-art oversampling algorithms significantly in terms of G-mean and F1-score, presenting a data distribution-focused solution for addressing imbalanced data challenges. [ABSTRACT FROM AUTHOR]

The key innovation lies in the incorporation of an adaptive cluster sampling strategy and a randomized response model based on the Poisson distribution. This integration aims to overcome shortcomings inherent in conventional models, providing a more robust framework for research area. In this paper, an adaptive cluster sampling randomized response model with Poisson distribution using a randomized response strategy was proposed. The proposed cluster randomized response model has improved efficiency and a large gain in precision. Conditions were obtained under which the proposed model is more efficient than the existing models. To validate the effectiveness of our approach, numerical computations were conducted, offering concrete illustrations of the model's performance. The results underscore the significant gains in efficiency and precision achieved by the proposed adaptive cluster sampling randomized response model. [ABSTRACT FROM AUTHOR]

This paper deals with making inference on parameters of a two‐level model matching the design hierarchy of a two‐stage sample. In a pioneering paper, Scott and Smith (Journal of the American Statistical Association, 1969, 64, 830–840) proposed a Bayesian model based or prediction approach to estimating a finite population mean under two‐stage cluster sampling. We provide a brief account of their pioneering work. We review two methods for the analysis of two‐level models based on matching two‐stage samples. Those methods are based on pseudo maximum likelihood and pseudo composite likelihood taking account of design weights. We then propose a new method for analysis of two‐level models based on a normal approximation to the estimated cluster effects and taking account of design weights. This method does not require cluster sizes to be constants or unrelated to cluster effects. We evaluate the relative performance of the three methods in a simulation study. Finally, we apply the methods to real data obtained from 2011 Nepal Demographic and Health Survey (NDHS). [ABSTRACT FROM AUTHOR]

Because the number of clustering cores needs to be set before implementing the K-means algorithm, this type of algorithm often fails in applications with increasing data and changing distribution characteristics. This paper proposes an evolutionary algorithm DCC, which can dynamically adjust the number of clustering cores with data change. DCC algorithm uses the Gaussian function as the activation function of each core. Each clustering core can adjust its center vector and coverage based on the response to the input data and its memory state to better fit the sample clusters in the space. The DCC algorithm model can evolve from 0. After each new sample is added, the winning dynamic core can be adjusted or split by competitive learning, so that the number of clustering cores of the algorithm always maintains a better adaptation relationship with the existing data. Furthermore, because its clustering core can split, it can subdivide the densely distributed data clusters. Finally, detailed experimental results show that the evolutionary clustering algorithm DCC based on the dynamic core method has excellent clustering performance and strong robustness. [ABSTRACT FROM AUTHOR]

Not only do cluster randomized trials require a larger sample size than individually randomized trials, they also face many additional complexities. The potential for contamination is the most commonly used justification for using cluster randomization, but the risk of contamination should be carefully weighed against the more serious problem of questionable scientific validity in settings with post-randomization identification or recruitment of participants unblinded to the treatment allocation. In this paper we provide some simple guidelines to help researchers conduct cluster trials in a way that minimizes potential biases and maximizes statistical efficiency. The overarching theme of this guidance is that methods that apply to individually randomized trials rarely apply to cluster randomized trials. We recommend that cluster randomization be only used when necessary—balancing the benefits of cluster randomization with its increased risks of bias and increased sample size. Researchers should also randomize at the lowest possible level—balancing the risks of contamination with ensuring an adequate number of randomization units—as well as exploring other options for statistically efficient designs. Clustering should always be allowed for in the sample size calculation; and the use of restricted randomization (and adjustment in the analysis for covariates used in the randomization) should be considered. Where possible, participants should be recruited before randomizing clusters and, when recruiting (or identifying) participants post-randomization, recruiters should be masked to the allocation. In the analysis, the target of inference should align with the research question, and adjustment for clustering and small sample corrections should be used when the trial includes less than about 40 clusters. [ABSTRACT FROM AUTHOR]

Screen time and the use of social networks is the most frequent form of leisure time occupation and socializing for adolescents. The present study is aimed at understanding and characterizing, from an ecological perspective, what distinguishes healthy and less healthy or even dependent use of social media in young people and the influence on adolescents' well-being. This paper is based on the Health Behaviour in School-aged Children (HBSC) from 2022, a survey carried out every 4 years, in collaboration with the World Health Organisation (WHO), following an international protocol. A total of 7643 students from the 6th, 8th, 10th, and 12th school grades responded, of which 53.9% were female, with an average age of 15.05 (SD = 2.36). The sample is representative of the school grades under study. The results allow us to study and identify similarities and differences between three groups related to the level of healthiness in the use of social networks and its relations to well-being from an ecological perspective. What distinguishes adolescents with less healthy use of social networks is that they are more often female, older, have more self-injurious behaviour, higher alcohol consumption, and a worse relationship with teachers. The adolescents with the highest level of dependence on social networks are those who have a higher perception of lack of safety at school and in their area of residence, as well as a higher use of screen time as a leisure activity. The well-being of adolescents using social media in a healthy way is explained by fewer psychological symptoms, better stress management strategies, better body awareness, more physical activity, less time online with friends, and better relationships with family and teachers. Technologies and social networks are important for the well-being of adolescents; it is essential to promote a healthy, critical and balanced use with other "screen-free" activities and to promote socio-emotional skills, a lack of which seems to be one of the biggest risk factors associated with the healthy use of technologies. [ABSTRACT FROM AUTHOR]

The available surface defect detection methods for disposable wooden spoons still involve screening with the naked eye. This detection method is not only inefficient but also accompanied by problems such as false detection and missed detection. Therefore, this paper proposes a detection method based on an improved YOLOv5 network model (YOLOv5-TSPP). This method uses the K-Means ++ algorithm to cluster the target samples in the data set to obtain anchor frames that are more in line with different target scales. The Coordinate Attention module is added to the backbone network of the YOLOv5 network model to improve the feature extraction ability of the model. A new SPP module is added to the backbone network to increase the important features in the receptive field extraction network to improve the detection accuracy of small targets. The experimental results show: The YOLOv5-TSPP algorithm has better detection performance and the mAP of defect detection reaches 80.3%, which is 9.2% higher than that of the YOLOv5 algorithm. Among them, the detection accuracy of black knot defect reached 98.6%, the detection accuracy of back crack defect reached 92.1%, and the detection accuracy of mineral line defect reached 92.3%. [ABSTRACT FROM AUTHOR]

Two‐stage randomized experiments become an increasingly popular experimental design for causal inference when the outcome of one unit may be affected by the treatment assignments of other units in the same cluster. In this paper, we provide a methodological framework for general tools of statistical inference and power analysis for two‐stage randomized experiments. Under the randomization‐based framework, we consider the estimation of a new direct effect of interest as well as the average direct and spillover effects studied in the literature. We provide unbiased estimators of these causal quantities and their conservative variance estimators in a general setting. Using these results, we then develop hypothesis testing procedures and derive sample size formulas. We theoretically compare the two‐stage randomized design with the completely randomized and cluster randomized designs, which represent two limiting designs. Finally, we conduct simulation studies to evaluate the empirical performance of our sample size formulas. For empirical illustration, the proposed methodology is applied to the randomized evaluation of the Indian National Health Insurance Program. An open‐source software package is available for implementing the proposed methodology. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Computer Engineering & Applications is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

It is thought to be an effective technique to handle the problem of educational data explosion and lack of information by identifying potential relationships between data and directing decision-makers through the extraction, transformation, analysis, and modeling of educational data. Based on this, this research constructs a data analysis model for education evaluation using the K-means clustering technique in DM. The weight of each index of students' comprehensive quality is calculated using AHP, and the value of the weight is used to determine whether the index is the important feature of analysis system mining. Improved sampling technology is used to deal with the representation of large-scale data sets; a sample partition clustering technique is proposed as a general framework. The best accuracy of this method, according to experimental data, is 95.6 percent, which is 12.1 percent greater than Mi cluster algorithm and 6.8 percent higher than DRCluster algorithm. The K-means clustering analysis technology is used to analyze students' comprehensive evaluation data in this paper, with the goal of determining the regularity of data implication, accurately diagnosing learning problems, and providing the foundation for developing effective student management strategies. [ABSTRACT FROM AUTHOR]

The aim of this paper is to clarify the difference of spatial closeness between natural enemies and Frankliniella intonsa, and to provide a scientific basis for rational protection and utilization of natural enemies. This paper applied the block‐sample square difference analysis method, Gray correlation degree method, the aggregation intensity index method and ρ index method to study the difference in the closeness of the spatial relationship between F. intonsa and natural enemies when the number of F. intonsa and their natural enemies was the highest and the minimum area occupied by F. intonsa individuals and colonies in five tea plantations in Hefei, Anhui province in 2021 and four tea plantations in 2020. The results were as follows: 1) Tetragnatha squamataKarsch, Xysticus ephippiafus, Erigonidium graminicolum and Theridion Octomaculatum were the top four natural enemies closely related to F. intonsa in 2021. The top three natural enemies closely related to F. intonsa in 2020 were Tetragnatha squamata Karsch, Xysticus ephippiafus and Theridion Octomaculatum. Two years ago, Tetragnatha squamata Karsch and Xysticus ephippiafus were the same. 2) Cluster size did not change the distribution patterns of natural enemies and F. intonsa. 3) The aggregation of F. intonsa was caused by their own causes or some environmental factors, and the aggregation of natural enemies was caused by environmental factors. 4) In 2021, the minimum area occupied by individuals and groups of F. intonsa in tea gardens was 2m2, and in 2020, it was 8 m2. [ABSTRACT FROM AUTHOR]

Cluster randomized experiments and estimator comparisons are well-documented topics. In this paper, using the datasets of the popular vote in the presidential elections of the United States of America (2012, 2016, 2020), we evaluate the properties (SE, MSE) of three cluster sampling estimators: Ratio estimator, Horvitz–Thompson estimator and the linear regression estimator. While both the Ratio and Horvitz–Thompson estimators are widely used in cluster analysis, we propose a linear regression estimator defined for unequal cluster sizes, which, in many scenarios, performs better than the other two. The main objective of this paper is twofold. Firstly, to indicate which estimator is most suited for predicting the outcome of the popular vote in the United States of America. We do so by applying the single-stage cluster sampling technique to our data. In the first partition, we use the 50 states plus the District of Columbia as primary sampling units, whereas in the second one, we use 3112 counties instead. Secondly, based on the results of the aforementioned procedure, we estimate the number of clusters in a sample for a set standard error while also considering the diminishing returns from increasing the number of clusters in the sample. The linear regression estimator is best in the majority of the examined cases. This type of comparison can also be used for the estimation of any other country’s elections if prior voting results are available.

AbstractThe authors identity the relationship between the positive and negative aspects of social media and the ideal belief learning and behavior of university students. The cluster sampling method was adopted in the paper, including Guangdong, Shandong, Henan, Sichuan, and Jiangsu provinces. A total of 1014 questionnaires were distributed to a purposive sample of university students between the ages of 16 and 35. The authors applied the uses and gratifications theory to study students’ social media behavior. This study identified 18 positive and negative effects of social media. Noteworthy positive outcomes attributed to social media in fostering ideals and beliefs encompass heightened awareness, enhanced communication facilitation, convenient connectivity, reduced expenses on educational materials, improved social and communication proficiencies, as well as diminished stress levels. The negative effects of new media and the Internet include a lack of critical thinking, a waste of time, dysgraphia, disrupted connection to learning, students’ laziness, and health risks. [ABSTRACT FROM AUTHOR]

Context. The Li abundance observed in pre-main sequence and main sequence late-type stars is strongly age-dependent, but also shows a complex pattern depending on several parameters, such as rotation, chromospheric activity, and metallicity. The best way to calibrate these effects, and with the aim of studying Li as an age indicator for FGK stars, is to calibrate coeval groups of stars, such as open clusters (OCs) and associations. Aims. We present a considerable target sample of 42 OCs and associations – with an age range from 1 Myr to 5 Gyr – observed within the Gaia-ESO survey (GES), and using the latest data provided by GES iDR6 and the most recent release of Gaia that was then available, EDR3. As part of this study, we update and improve the membership analysis for all 20 OCs presented in our previous article. Methods. We perform detailed membership analyses for all target clusters to identify likely candidates, using all available parameters provided by GES, complemented with detailed bibliographical searches, and based on numerous criteria: from radial velocity distributions, to the astrometry (proper motions and parallaxes) and photometry provided by Gaia, to gravity indicators (log g and the γ index), [Fe/H] metallicity, and Li content in diagrams of (Li equivalent widths) EW(Li) versus Teff. Results. We obtain updated lists of cluster members for the whole target sample, as well as a selection of Li-rich giant contaminants obtained as an additional result of the membership process. Each selection of cluster candidates was thoroughly contrasted with numerous existing membership studies using data from Gaia to ensure the most robust results. Conclusions. These final cluster selections will be used in the third and last paper of this series, which reports the results of a comparative study characterising the observable Li dispersion in each cluster and analysing its dependence on several parameters, allowing us to calibrate a Li–age relation and obtain a series of empirical Li envelopes for key ages in our sample. [ABSTRACT FROM AUTHOR]

With clustered data, such as where students are nested within schools or employees are nested within organizations, it is often of interest to estimate and compare associations among variables separately for each level. While researchers routinely estimate between-cluster effects using the sample cluster means of a predictor, previous research has shown that such practice leads to biased estimates of coefficients at the between level, and recent research has recommended the use of latent cluster means with the multilevel structural equation modeling framework. However, the latent cluster mean approach may not always be the best choice as it (a) relies on the assumption that the population cluster sizes are close to infinite, (b) requires a relatively large number of clusters, and (c) is currently only implemented in specialized software such as Mplus. In this paper, we show how using empirical Bayes estimates of the cluster means can also lead to consistent estimates of between-level coefficients, and illustrate how the empirical Bayes estimate can incorporate finite population corrections when information on population cluster sizes is available. Through a series of Monte Carlo simulation studies, we show that the empirical Bayes cluster-mean approach performs similarly to the latent cluster mean approach for estimating the between-cluster coefficients in most conditions when the infinite-population assumption holds, and applying the finite population correction provides reasonable point and interval estimates when the population is finite. The performance of EBM can be further improved with restricted maximum likelihood estimation and likelihood-based confidence intervals. We also provide an R function that implements the empirical Bayes cluster-mean approach, and illustrate it using data from the classic High School and Beyond Study. [ABSTRACT FROM AUTHOR]

There have been reported association of oral health disorders with burnout, stress, and mental health. Arguably, with these reported associations, and the current prevalence of burnout amongst Nigerian doctors, exploring the role of burnout on oral health amongst Nigerian doctors is timely. This study aims to determine the relationship between burnout and oral health-related quality of life amongst Early Career Doctors (ECDs) in Nigeria, while also identifying the role other possible predictors plays in this relationship. This was a cross-sectional study conducted amongst Nigerian ECDs as part of Challenges of Residency Training in Nigeria (CHARTING) II project. A total of 632 ECDs were recruited across thirty-one tertiary hospitals in the 6 geopolitical zones of the country using a multistage cluster sampling technique. A self-administered paper-based semi-structured questionnaire was given to each participant that consented. The tools used to assess burnout and Oral health-related quality of life (OHRQoL) were Copenhagen Burnout Inventory (CBI) and Oral Health Impact Profile (OHIP-14) respectively. Independent samples T-test, ANOVA and Multiple linear regression were used to draw inferences from the data collected. Overall mean OHIP-14 score of all participants was 11.12 (±9.23). The scores for the 3 dimensions of burnout were below 50% with CBI-Personal Burnout having the highest score of 49.96 (±19.15). Significant positive correlations (p < 0.001) were found between OHIP-14 and all the dimensions of burnout, as the burnout scores were increasing, there was a corresponding increase in the OHIP scores thus poorer OHRQoL. The regression model shows that the predictors of OHIP were CBI-PB (p = 0.003), use of fluoride paste (p = 0.039), use of tobacco (p = 0.005) and being a denture user (p = 0.047). This study shows a positive correlation between burnout and OHIP of ECDs. We found that as burnout was increasing, OHIP increased thus implying poorer oral health related quality of life amongst ECDs. The use of fluoride toothpaste, tobacco and denture are other factors we found that could affect the OHIP of ECDs. [ABSTRACT FROM AUTHOR]

Segregation in the housing sector has exposed urban dwellers to series of unfavourable conditions in most cities of developing countries, which are as a result several factors. This paper is aimed at examining the factors influencing segregation in the housing neighbourhoods of Minna, Nigeria. Two-stage cluster sampling methods were adopted in the selection of neighbourhoods and households across the entire sample frame in the administration of 374 questionnaires across the 25 clustered neighbourhoods. Chi-square and Relative Importance Index (RII) as well as mean score to determine neighbourhood geographical segregation indicators pattern across all the 25 neighbourhoods in the study area. The study established that there is a significant difference between low, medium, and high density neighbourhoods in terms segregation factors in the study area. It was therefore suggested that a good urban governance structure should be put in place that will discourage class divisions among spatial entities of the city. [ABSTRACT FROM AUTHOR]

Over the last four decades, the cluster regression analysis in a finite population (FP) setup for an exponential family such as linear or binary data was done by using a two-stage cluster sample chosen from the FP but by treating the sample as though it is a single-stage cluster sample from a super-population (SP) which contains the FP as a hypothetical sample. Because the responses within a cluster in the FP are correlated, the aforementioned sample mis-specification makes the sample-based so-called GLS (generalized least square) estimators design biased and inconsistent. In this paper, we demonstrate for the exponential family data how to avoid the sampling mis-specification and accommodate the cluster correlations to obtain unbiased and consistent estimates for the FP parameters. The asymptotic normality of the regression estimators is also given for the construction of confidence intervals when needed. [ABSTRACT FROM AUTHOR]

This research note discusses the benefits of guided tours as an online data collection strategy. These benefits include an opportunity to follow-up on questions and seek detailed responses from participants, clarify certain misconceptions about the participants and/or phenomena under study, mitigate bias in data, get an updated record of participants' online practices, and create a relationship of trust and rapport with the participants. Guided tours can also be useful for participants as it gives them a chance to use the online space regularly, and also provides them with an opportunity to reflect on their online practices frequently. Keeping in view the benefits of guided tours, qualitative researchers are encouraged to employ this data collection strategy in their online research for securing in-depth and additional information regarding participants' use of a particular online site. Further, this short paper recommends that online qualitative researchers utilize guided tours to seek additional and detailed information about their participants' practices and/or experiences in any online space. [ABSTRACT FROM AUTHOR]

There are a large number of imbalanced classification problems in the real world. Due to the imbalance in the amount of data and the complex nature of the distribution, the minority class samples are difficult to be classified correctly. Oversampling techniques balance the data set by generating minority class samples; however, current clustering-based oversampling techniques are limited by hyperparameters, sample selection, and other issues that affect the final classification performance. In this paper, we propose an oversampling algorithm based on natural neighbor and density peaks clustering (ND-S). ND-S is divided into three steps. Firstly, the natural neighbor algorithm is used to find and filter noises and outliers. Secondly, the density peaks clustering is improved by natural neighbor-based nonparametric adaptive, which clusters all samples and leaves the clusters that meet the conditions. Finally, sampling weights are assigned to each cluster, and the minority class of samples suitable for oversampling is selected for synthetic minority oversampling (SMOTE) by calculating the local sparsity of the samples. Experiments on 18 imbalanced data sets show that ND-S is effective for the imbalanced classification problem, and its classification performance is generally better than other 8 comparison algorithms. [ABSTRACT FROM AUTHOR]

Person reidentification (re-ID) is an important topic in computer vision. This paper studies an unsupervised approach to re-ID, which does not require any labeled information and is thus possible to deploy in real-world scenarios. State-of-the-art unsupervised re-ID methods usually use a memory bank to store the instance feature vectors, generate pseudolabels with a clustering algorithm, and compare the query instances to the centroid of the clusters for contrastive learning. However, because hard negative or noisy samples exist, the centroid generated by unsupervised learning may not be a perfect prototype. Forcing the wrong images to get closer to the centroid would result in accumulated errors and deteriorated overfitting. To solve this problem, we propose a quantitative random selection strategy to form the cluster feature representation. Specifically, in each iteration, the cluster algorithm executes on instance-level feature vectors to generate pseudolabels. Then, we shuffle all the instance vectors belonging to the same cluster and select samples within the same cluster in a certain proportion to form the cluster-level memory. During network training, the query instances are used to update the cluster-level memory for contrastive learning. Extensive experiments show that our proposed method produces state-of-the-art performance in unsupervised person re-ID tasks. [ABSTRACT FROM AUTHOR]

Copyright of Pamukkale University Journal of Social Sciences Institute / Pamukkale Üniversitesi Sosyal Bilimler Enstitüsü Dergisi is the property of Pamukkale University, Social Sciences Institute and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)