Cluster-based Sampling Research Articles

Fostering resilience and post-traumatic growth in overseas Chinese left-behind children: The role of autonomy, competence, and relatedness.

This study examines the impact of parental migration on the psychological well-being and development of left-behind children (LBCs) in Zhejiang, China, within the broader context of the country's rural transformations and urban migration. It investigates how intellectual and relational engagement (RE), autonomy (AUT), competence (COM), and relatedness (RES) contribute to resilience (REL) and post-traumatic growth (PTG) in these children, reflecting on the shift from viewing parental separation merely as a source of trauma to recognizing its potential to foster significant personal growth. Utilizing a cross-sectional design, the research was conducted in April and May 2023 with 1348 LBCs from a total sample of 4049 students inZhejiang. A two-step random, stratified, cluster-based sampling strategy was employed, and structural equation modeling was used to examine the hypothesized relationships among the constructs. The statistical analysis demonstrated significant positive effects of intellectual engagement (IE), AUT, COM, and RE on both REL and PTG (p<.05 for all). IE strongly correlated with AUT (r=.68, p<.001) and COM (r=.71, p<.001), enhancing REL and facilitating PTG. Additionally, the presence of secure and consistent relationships was identified as crucial for maintaining psychological well-being, with high correlation coefficients (r>.60) underscoring their importance. Notably, REL was found to moderate the relationships among RES, COM, and PTG, highlighting its critical role in the psychological adaptation of left-behind children. The study underscores the importance of nurturing intellectual and REs, AUT, and COM to enhance psychological REL and well-being among LBAs. These elements are crucial for supporting the mental health and developmental needs of children facing the challenges of parental migration. The findings advocate for targeted interventions that can address the unique needs of this vulnerable population, emphasizing the potential for growth and adaptation despite adversities.

Withdrawal Notice: Cluster-based Sampling for Fake News on Healthcare During COVID-19 Infodemic

This article has been withdrawn from the journal Recent Advances in Electrical and Electronic Engineering. Bentham Science apologizes to the readers of the journal for any inconvenience this may have caused. The Bentham Editorial Policy on Article Withdrawal can be found at https://benthamscience.com/editorial-policies-main.php. BENTHAM SCIENCE DISCLAIMER: It is a condition of publication that manuscripts submitted to this journal have not been published and will not be simultaneously submitted or published elsewhere. Furthermore, any data, illustration, structure or table that has been published elsewhere must be reported, and copyright permission for reproduction must be obtained. Plagiarism is strictly forbidden, and by submitting the article for publication the authors agree that the publishers have the legal right to take appropriate action against the authors, if plagiarism or fabricated information is discovered. By submitting a manuscript, the authors agree that the copyright of their article is transferred to the publishers if and when the article is accepted for publication.

Cluster-Based Sampling in Hindsight Experience Replay for Robotic Tasks (Student Abstract)

In multi-goal reinforcement learning with a sparse binary reward, training agents is particularly challenging, due to a lack of successful experiences. To solve this problem, hindsight experience replay (HER) generates successful experiences even from unsuccessful ones. However, generating successful experiences from uniformly sampled ones is not an efficient process. In this paper, the impact of exploiting the property of achieved goals in generating successful experiences is investigated and a novel cluster-based sampling strategy is proposed. The proposed sampling strategy groups episodes with different achieved goals by using a cluster model and samples experiences in the manner of HER to create the training batch. The proposed method is validated by experiments with three robotic control tasks of the OpenAI Gym. The results of experiments demonstrate that the proposed method is substantially sample efficient and achieves better performance than baseline approaches.

Comparison of Cluster-Based Sampling Approaches for Imbalanced Data of Crashes Involving Large Trucks

Severe and fatal crashes involving large trucks result in significant social and economic losses for human society. Unfortunately, the notably low proportion of severe and fatal injury crashes involving large trucks creates an imbalance in crash data. Models trained on imbalanced crash data are likely to produce erroneous results. Therefore, there is a need to explore novel sampling approaches for imbalanced crash data, and it is crucial to determine the appropriate combination of a machine learning model, sampling approach, and ratio. This study introduces a novel cluster-based under-sampling technique, utilizing the k-prototypes clustering algorithm. After initial cluster-based under-sampling, the consolidated cluster-based under-sampled data set was further resampled using three different sampling approaches (i.e., adaptive synthetic sampling (ADASYN), NearMiss-2, and the synthetic minority oversampling technique + Tomek links (SMOTETomek)). Later, four machine learning models (logistic regression (LR), random forest (RF), gradient-boosted decision trees (GBDT), and the multi-layer perceptron (MLP) neural network) were trained and evaluated using the geometric mean (G-Mean) and area under the receiver operating characteristic curve (AUC) scores. The findings suggest that cluster-based under-sampling coupled with the investigated sampling approaches improve the performance of the machine learning models developed on crash data significantly. In addition, the GBDT model combined with ADASYN or SMOTETomek is likely to yield better predictions than any model combined with NearMiss-2. Regarding changes in sampling ratios, increasing the sampling ratio with ADASYN and SMOTETomek is likely to improve the performance of models up to a certain level, whereas with NearMiss-2, performance is likely to drop significantly beyond a specific point. These findings provide valuable insights for selecting optimal strategies for treating the class imbalance issue in crash data.

Assessment of Knowledge Regarding Postnatal Care among Men in Selected Rural Areas of Himachal Pradesh

Background: WHO defined health as the state of complete physical, mental, and social well-being and not merely an absence of disease or infirmity, but it includes ability to lead socially and economically productive life.1 Being a man or women has a significant impact on health, as a result of both biological as well as gender related differences. This study assesses the knowledge about the antenatal care among married and unmarried men of 18-35 years age in selected rural areas of Himachal Pradesh. Methods: The study was an unmatched case and control design where married men were considered as a case and unmarried men as comparator group. This study was a part of large study for which sample size of 268 men in each group was calculated assuming prevalence of female contraceptives 70% in unmarried men and 85% in married men. This was carried out by cluster-based sampling in villages of Nagrota Bagwan and Shahpur health blocks of DRPGMC Tanda at Kangra. Results: The 93.3% participants among case group are aware of anaemia in pregnancy while only 64.9% among control group are aware of anaemia in pregnancy. The knowledge about all three causes of anaemia was statistically more in case group (61.9%) as compared to control group (38.8%). Assessment about high-risk pregnancy, about 72.4% participants among case group were aware about the risk factors in pregnancy as compared to 43.3% in control group with statistically significant differences (P=0.000). All the causes of high-risk pregnancy were reported by the participants of case group (44.0% vs 29.9%) with statistically significant differences. Conclusions: Married men had significantly more knowledge about various components of antenatal care. The awareness approach for unmarried men needs to be designed regarding antenatal care.

Assessment of Knowledge Regarding Antenatal Care among Men in Selected Rural Areas of Himachal Pradesh

Background: WHO defined health as the state of complete physical, mental, and social well-being and not merely an absence of disease or infirmity, but it includes ability to lead socially and economically productive life.1 Being a man or women has a significant impact on health, as a result of both biological as well as gender related differences. Antenatal care (ANC) is most important aspect of women‘s reproductive health. The antenatal care is the care provided by skilled health care professionals to pregnant women and adolescent girls to ensure best health conditions for mother and baby during pregnancy. It reduces maternal and perinatal morbidity and mortality both directly, through detections and indirectly, through the identification of women and girls at increased risk of developing complications during labour, delivery and ensuring appropriate level of care.1 As per NFHS- 4, the 21% of pregnant women utilised full ANC, ranging from 2.3–65.9% across states. Overall, 51.6% had 4 or more ANC visits, 30.8% consumed IFA for at least 100 days and 91.1% had one or more doses of tetanus toxoid. Full ANC utilisation was inequitable across place of residence, caste, and maternal education. Registration of pregnancy, utilisation of government‘s Integrated Child Development Services (ICDS) and health insurance coverage were associated with higher odds of full ANC utilisation. Lower maternal education, lower wealth quintile, lack of father‘s participation during antenatal visits, higher birth order, teenage and unintended pregnancy were associated with lower odds of full ANC utilisation. This study assesses the knowledge about the antenatal care among married and unmarried men of 18-35 years age in selected rural areas of Himachal Pradesh. Methods: The study was an unmatched case and control design where married men were considered as a case and unmarried men as comparator group. This study was a part of large study for which sample size of 268 men in each group was calculated assuming prevalence of female contraceptives 70% in unmarried men and 85% in married men. This was carried out by cluster-based sampling in villages of Nagrota Bagwan and Shahpur health blocks of DRPGMC Tanda at Kangra. Results: The 93.3% participants among case group are aware of anaemia in pregnancy while only 64.9% among control group are aware of anaemia in pregnancy. The knowledge about all three causes of anaemia was statistically more in case group (61.9%) as compared to control group (38.8%). Assessment about high-risk pregnancy, about 72.4% participants among case group were aware about the risk factors in pregnancy as compared to 43.3% in control group with statistically significant differences (P=0.000). All the causes of high-risk pregnancy were reported by the participants of case group (44.0% vs 29.9%) with statistically significant differences. Conclusions: Married men had significantly more knowledge about various components of antenatal care. The awareness approach for unmarried men needs to be designed regarding antenatal care.

An end-to-end weakly supervised learning framework for cancer subtype classification using histopathological slides

AI-powered analysis of histopathology data has become an invaluable assistant for pathologists due to its efficiency and accuracy. However, existing deep learning methods still face some challenges in specifying cancer subtypes. For example, the ultra-high resolution of histopathological slides generally contains numerous redundant features, which are not useful for cancer subtype classification and thus lead to considerable computational costs. Moreover, the lack of manual annotations of disease-specific regions (i.e., patch-level annotations) from experts makes it more difficult to learn such histological features with only slide-level labels. In this paper, we propose an end-to-end weakly supervised learning framework called EWSLF to address these issues. First, we employ a cluster-based sampling strategy to refine the histological features for further training, which can improve classification accuracy and reduce computational cost. Second, we employ a multi-branch attention mechanism to produce patch-level pseudo-labels and aggregate the patch features into slide-level features, which can complement the missing patch-level labels from experts. Experimental results on both public and in-house datasets demonstrate the superiority and credible results of our model compared with the state-of-the-art methods for cancer subtype classification. Code: https://github.com/hongren21/ewslf.

Self-Supervised Local Topology Representation for Random Cluster Matching

This letter aims to learn a global representation for each point in a random cluster using only purely local geometric or topological information. Based on this, distributed tags for indoor positioning break the atomicity of tags and make deployment more arbitrary. It also allows NP-hard matches to be quickly estimated with only one local observation. The novel self-supervised topological representation learning method only takes local point clusters as input and utilizes the proposed cluster-based sampling, training, and loss functions to form global self-comparison. The training samples are generated in real-time virtually, and there are few matching errors after being transferred to practice. The compact backbone network directly processes the coordinates of points and abandons the iterative optimization commonly used in matching. Moreover, it uses the representation to measure similarity directly, and the inference speed reaches the millisecond level. In the actual and virtual experiments, the local point clusters are surprisingly accurately matched to the random global ones. The localization based on this is also verified, and the relevant results prove the effectiveness of the proposed method.

Practical strategies for operationalizing optimal allocationin stratified cluster-based outcome-dependent sampling designs.

Cluster-based outcome-dependent sampling (ODS) has the potential to yield efficiency gains when the outcome of interest is relatively rare, and resource constraints allow only a certain number of clusters to be visited for data collection. Previous research has shown that when the intended analysis is inverse-probability weighted generalized estimating equations, and the number of clusters that can be sampled is fixed, optimal allocation of the (cluster-level) sample size across strata defined by auxiliary variables readily available at the design stage has the potential to increase efficiency in the estimation of the parameter(s) of interest. In such a setting, the optimal allocation formulae depend on quantities that are unknown in practice, currently making such designs difficult to implement. In this paper, we consider a two-wave adaptive sampling approach, in which data is collected from a first wave sample, and subsequently used to compute the optimal second wave stratum-specific sample sizes. We consider two strategies for estimating the necessary components using the first wave data: an inverse-probability weighting (IPW) approach and a multiple imputation (MI) approach. In a comprehensive simulation study, we show that the adaptive sampling approach performs well, and that the MI approach yields designs that are very near-optimal, regardless of the covariate type. The IPW approach, on the other hand, has mixed results. Finally, we illustrate the proposed adaptive sampling procedures with data on maternal characteristics and birth outcomes among women enrolled in the Safer Deliveries program in Zanzibar, Tanzania.

Cluster-based stratified sampling for fast reliability evaluation of composite power systems based on sequential Monte Carlo simulation

An essential aspect in the operation and expansion of the electricity infrastructure is to ensure a reliable power supply, despite uncertainty (mainly caused by random failures, variability of renewable sources, and load growth). Reliability evaluation of composite power systems is a valuable tool for identifying possible deficiencies in operation. Power systems operation is becoming more variable and stochastic. Consequently, there is an urgent need to update the tools used to analyze their reliability. This article presents a novel method based on cluster-based stratified sampling (CBSS) and sequential Monte Carlo simulations (SMCS) to improve the reliability evaluation of composite power systems. Here, clustering algorithms are applied to reduce the number of observations required by traditional SMCS. The proposed approach is applied to RTS-79, RTS-96, and RTS-GMLC electrical networks to verify its accuracy and speed. The results obtained demonstrate that CBSS increase calculation speed in highly reliable networks, while preserving information on the probability distribution of the reliability indices.

The Impact of Information and Communication Technology on the User Services Division of the Forde Library of University of the Southern Caribbean

This study seeks to assess the user frequency and impact of ICT on the performance of the User Services Division, with specific insight on the Circulation, Reference and Instruction Units of the Forde Library. The impact of ICT was measured using cluster-based sampling given to 340 undergraduate students, with a return rate of 62% - 211 students. The analysis revealed a significant lack of awareness and usage of the ICT offerings. Exceptions to low usage and awareness were the OPAC and the Library’s Website. The main factors that impacted the low awareness and usage were limited marketing, no knowledge of ICT offerings, no need for the offerings and complications experienced while using the offerings. Overall, the benefits of the ICT offerings were not recognized by the sample size and as such, lower scores of relevancy, ease of use and user-friendliness were recorded. Recommendations focused on marketing, training, and interdepartmental collaboration.

Gender modified association of oral health indicators with oral health-related quality of life among Korean elders

ObjectiveTo evaluate the association between oral health-related quality of life (OHRQoL) and oral health indicators including dental status, total occlusion force (TOF), number of natural and rehabilitated teeth (NRT), number of natural teeth (NT), and to explore the effect modification on the association by gender among Korean elders.MethodsA total of 675 participants aged 65 or above recruited by a cluster-based stratified random sampling were included in this cross-sectional study. The 14-items Korean version of the Oral Health Impact Profile (OHIP) was used to measure OHRQoL. The responses about OHIP were dichotomized by the cut-off point of 'fairly often' to determine the ‘poor’ versus ‘fair’ OHRQoL. Age, gender, education level, alcohol drinking, smoking, metabolic syndrome, frailty, and periodontitis were considered as confounders. Multiple multivariable logistic regression analyses were applied to assess the adjusted association between oral health indicators and OHRQoL. Gender stratified analysis was also applied to explore the effect modification of the association.ResultsThe prevalence of poor OHRQoL was 43.0%, which was higher in women, less-educated elders, alcohol non-drinkers and frailty elders (p < 0.05). Elders with poor OHRQoL also showed lower values of oral health indicators than elders with fair OHRQoL (p < 0.05). Those with NRT ≤ 24, NT ≤ 14, and TOF < 330 N increased the risk of poor OHRQoL by 2.3 times (OR = 2.26, confidence interval [CI] 1.54–3.31), 1.5 times (OR = 1.45, CI 1.02–2.07), and 1.5 times (OR = 1.47, CI 1.06–2.04), respectively. In women, the association of NRT ≤ 24 with poor OHRQoL increased from OR of 2.3 to OR of 2.4, while, in men, the association of TOF < 330 N with poor OHRQoL increased from OR of 1.5 to OR of 3.2.ConclusionOral health indicators consisting of TOF, NRT, and NT were independently associated with poor OHRQoL among Korean elders. Gender modified the association of TOF and NRT. Preventive and/or curative management for keeping natural teeth and the rehabilitation of missing teeth to recover the occlusal force may be essential for reducing poor OHRQoL.

Ligand-based approaches to activity prediction for the early stage of structure-activity-relationship progression.

The retrospective evaluation of virtual screening approaches and activity prediction models are important for methodological development. However, for fair comparison, evaluation data sets must be carefully prepared. In this research, we compiled structure-activity-relationship matrix-based data sets for 15 biological targets along with many diverse inactive compounds, assuming the early stage of structure-activity-relationship progression. To use a large number of diverse inactive compounds and a limited number of active compounds, similarity profiles (SPs) are proposed as a set of molecular descriptors. Using these highly imbalanced data sets, we evaluated various approaches including SPs, under-sampling, support vector machine (SVM), and message passing neural networks. We found that for the under-sampling approaches, cluster-based sampling is better than random sampling. For virtual screening, SPs with inactive reference compounds and the under-sampling SVM also perform well. For classification, SPs with many inactive references performed as well as the under-sampling SVM trained on a balanced data set. Although the performance of SPs and the under-sampling SVM were comparable, SPs with many inactive references were preferable for selecting structurally distinct compounds from the active training compounds.

Accelerating Energy-Economic Simulation Models via Machine Learning-Based Emulation and Time Series Aggregation

Energy-economic simulation models with high levels of detail, high time resolutions, or large populations (e.g., distribution networks, households, electric vehicles, energy communities) are often limited due to their computational complexity. This paper introduces a novel methodology, combining cluster-based time series aggregation and sampling methods, to efficiently emulate simulation models using machine learning and significantly reduce both simulation and training time. Machine learning-based emulation models require sufficient and high-quality data to generalize the dataset. Since simulations are computationally complex, their maximum number is limited. Sampling methods come into play when selecting the best parameters for a limited number of simulations ex ante. This paper introduces and compares multiple sampling methods on three energy-economic datasets and shows their advantage over a simple random sampling for small sample-sizes. The results show that a k-means cluster sampling approach (based on unsupervised learning) and adaptive sampling (based on supervised learning) achieve the best results especially for small sample sizes. While a k-means cluster sampling is simple to implement, it is challenging to increase the sample sizes if the emulation model does not achieve sufficient accuracy. The iterative adaptive sampling is more complex during implementation, but can be re-applied until a certain accuracy threshold is met. Emulation is then applied on a case study, emulating an energy-economic simulation framework for peer-to-peer pricing models in Germany. The evaluated pricing models are the “supply and demand ratio” (SDR) and “mid-market rate pricing” (MMR). A time series aggregation can reduce time series data of municipalities by 99.4% with less than 5% error for 98.2% (load) and 95.5% (generation) of all municipalities and hence decrease the simulation time needed to create sufficient training data. This paper combines time series aggregation and emulation in a novel approach and shows significant acceleration by up to 88.9% of the model’s initial runtime for the simulation of the entire population of around 12,000 municipalities. The time for re-calculating the population (e.g., for different scenarios or sensitivity analysis) can be increased by a factor of 1100 while still retaining high accuracy. The analysis of the simulation time shows that time series aggregation and emulation, considered individually, only bring minor improvements in the runtime but can, however, be combined effectively. This can significantly speed up both the simulation itself and the training of the emulation model and allows for flexible use, depending on the capabilities of the models and the practitioners. The results of the peer-to-peer pricing for approximately 12,000 German municipalities show great potential for energy communities. The mechanisms offer good incentives for the addition of necessary flexibility.

A Hybrid Sampling Approach for Imbalanced Binary and Multi-Class Data Using Clustering Analysis

Unequal data distribution among different classes usually cause a class imbalance problem. Due to the class imbalance, the classification models become biased toward the majority class and misclassify the minority class. Class imbalance issue becomes more complex when it occurs in multi-class data. The most common method to handle the class imbalance is data resampling that involves either over-sampling minority class instances or under-sampling majority class instances. In the case of under-sampling, there is a chance of losing some crucial information, whereas over-sampling can cause an overfitting problem. Therefore, we propose a novel Cluster-based Hybrid Sampling for Imbalance Data (CBHSID) strategy to address these issues. We calculate the mean of the data observations based on the number of classes. CBHSID uses the calculated mean as a threshold value to segregate majority and minority classes. We apply affinity propagation cluster analysis to each class to create sub-clusters. We calculate the distance of each data item of sub-cluster using centroid mean. We remove data observations that are away from the center of sub-cluster during under-sampling. On the other hand, during the over-sampling, we generate synthetic samples using data observations near to the center of sub-cluster. We compared our proposed approach with a few state-of-the-art data balancing methods on 12 binary and 4 multi-class benchmark datasets. Based on Geometric-Mean (G-Mean), Recall, and F1-score, our method outperformed the other compared methods on 14 datasets out of 16. We identified that CBHSID is suitable for addressing class imbalance issues in both binary and multi-class classifications. In the current state, we have only validated CBHSID on stationary data streams. Consequently, CBHSID can further be tested on non-stationary data streams in online learning environments.

Vulnerabilidades sociais e iniciação sexual entre 10 e 14 anos em estudantes do município do Rio de Janeiro, Brasil

Abstract This study aimed to estimate the prevalence of sexual initiation of pupils aged 10 to 14 who attended the second year of public and private high school in the IX Administrative Region of the city of Rio de Janeiro, Brazil and to identify the most vulnerable subgroups. The sample consisted of 694 pupils who were selected through cluster-based and stratified sampling by considering school type (public or private) and course type (daytime or evening). Information was collected by means of a structured self-administered questionnaire. The chi-square test (χ2) and 95% confidence intervals were used to assess the heterogeneity of proportions among subgroups. The prevalence of the event was 18.4%; it was higher in boys, in subgroups of greater social vulnerability, among those who hooked up/dated up to 14 years of age, in victims of sexual violence in affective-sexual relationships and in pupils showing health risk behaviors. The high rate of sexual initiation in early adolescence, especially in more vulnerable groups, shows that the situation must be understood and addressed by means of intersectoral public policies that take into account a social context of multiple needs rather than reproductive health alone.

A Hierarchical Clustering Method of SOM Based on DTW Distance for Variable-Length Seismic Waveform

In seismic facies analysis, waveform clustering of self-organizing map (SOM) usually classifies waveforms truncated by horizon with a fixed time window. However, the fixed time window is not an appropriate method for the situation in which the thickness of the target layer varies in the horizontal direction. In order to adapt SOM to this situation and simultaneously preserve the advantages of SOM, we replace the vanilla SOM measure with dynamic time warp (DTW) distance. Compared with Euclidean distance, DTW distance requires more computational amount about path searching and weight parameter updating in SOM. Therefore, considering the redundancy of seismic data, we introduce a hierarchical clustering strategy, which uses cluster-based stratified sampling and hierarchical mapping to reduce the computational cost of training and prediction. In the visualization section, by combining with hue, saturation, value (HSV) coloring and hierarchical mapping, the method can quickly display the lateral distribution of stratigraphic. The proposed method is validated by a mound shape model in a simulation test. Finally, the method was successfully applied to the field data from a reef-bank reservoir. The experimental and application results show that, when the top and bottom interfaces of the destination layer can be clearly identified, the proposed method can effectively cluster the waveforms with variable length, and the computational cost is acceptable.

Prevalence of Paediatric Surgical Conditions in Eastern Uganda: A Cross-Sectional Study

BackgroundThe role of surgery in global health has gained greater attention in recent years. Approximately 1.8 billion children below 15 years live in low- and middle-income countries (LMIC). Many surgical conditions affect children. Therefore, paediatric surgery requires specific emphasis. Left unattended, the consequences can be dire. Despite this, there is a paucity of data regarding prevalence of surgical conditions in children in LMIC. The present objective was to investigate the prevalence of paediatric surgical conditions in children in a defined geographical area in Eastern Uganda.MethodA cross-sectional study was carried out in the Iganga-Mayuge Health and Demographic Surveillance Site located in Eastern Uganda. Through a two-stage, cluster-based sampling process, 490 households from 49 villages were randomly selected, generating a study population of 1581 children. The children’s caregivers were interviewed, and the children were physically examined by two medical doctors to identify any surgical conditions.ResultsThe interview was performed with 1581 children, and 1054 were physically examined. Among these, the overall prevalence of any surgical condition was 16.0 per cent (n = 169). Of these, 39 per cent had an unmet surgical need (66 of 169). This is equivalent to a 6.3 per cent prevalence of current unmet surgical need. The most common groups of surgical condition were congenital anomalies and trauma-related conditions.ConclusionSurgical conditions in children are common in eastern Uganda. The unmet need for surgery is high. With a growing population, the need for paediatric surgical capacity will increase even further. The health care system must be reinforced to provide services for children with surgical conditions if United Nations Sustainability Development Goal 3 is to be achieved by 2030.

An Integrated District Mapping Strategy for Loiasis to Enable Safe Mass Treatment for Onchocerciasis in Gabon.

ABSTRACT.The lack of a WHO-recommended strategy for onchocerciasis treatment with ivermectin in hypo-endemic areas co-endemic with loiasis is an impediment to global onchocerciasis elimination. New loiasis diagnostics (LoaScope; Loa antibody rapid test) and risk prediction tools may enable safe mass treatment decisions in co-endemic areas. In 2017–2018, an integrated mapping strategy for onchocerciasis, lymphatic filariasis (LF), and loiasis, aimed at enabling safe ivermectin treatment decisions, was piloted in Gabon. Three ivermectin-naïve departments suspected to be hypo-endemic were selected and up to 100 adults per village across 30 villages in each of the three departments underwent testing for indicators of onchocerciasis, LF, and loiasis. An additional 67 communities in five adjoining departments were tested for loiasis to extend the prevalence and intensity predictions and possibly expand the boundaries of areas deemed safe for ivermectin treatment. Integrated testing in the three departments revealed within-department heterogeneity for all the three diseases, highlighting the value of a mapping approach that relies on cluster-based sampling rather than sentinel sites. These results suggest that safe mass treatment of onchocerciasis may be possible at the subdepartment level, even in departments where loiasis is present. Beyond valuable epidemiologic data, the study generated insight into the performance of various diagnostics and the feasibility of an integrated mapping approach utilizing new diagnostic and modeling tools. Further research should explore how programs can combine these diagnostic and risk prediction tools into a feasible programmatic strategy to enable safe treatment decisions where loiasis and onchocerciasis are co-endemic.

Audio based depression detection using Convolutional Autoencoder

Depression is a serious and common psychological disorder that requires early diagnosis and treatment. In severe episodes the condition may result in suicidal thoughts. Recently, the need for building an effective audio-based Automatic Depression Detection (ADD) system has sparked the interest of the research community. To date, most of the reported approaches to recognize depression rely on hand-crafted feature extraction for audio data representation. They combine wide variety of audio-related features to improve the classification performance. However, combining many hand-crafted features including relevant and less-relevant can enlarge the feature space which can lead to high-dimensionality issues as not all the features would carry significant information regarding depression. Having high number of features can make the pattern recognition more difficult and increase the risk of overfitting. To overcome these limitations, an audio-based framework of depression detection which includes an adaptation of a deep learning (DL) technique is proposed to automatically extract the highly relevant and compact feature set. This proposed framework uses an end-to-end Convolutional Neural Network-based Autoencoder (CNN AE) technique to learn the highly relevant and discriminative features from raw sequential audio data, and hence to detect depressed people more accurately. In addition, to address the sample imbalance problem we use a cluster-based sampling technique which highly reduces the risk of bias towards the major class (non-depressed). To evaluate the performance and effectiveness of the proposed pipeline, we perform the experiments on Distress Analysis Interview Corpus-Wizard of Oz (DAIC-WOZ) dataset and compare them with the hand-crafted feature extraction methods and other outstanding studies in this domain. The results show that proposed method outperforms other well-known audio-based ADD models with at least 7% improvement in F-measure for classifying depression.