Search

BACKGROUND: Food safety is pivotal for public welfare and directly impacts consumer health. Food safety sampling inspections (FSSIs) are essential in detecting unqualified food products and non‐compliant manufacturers, which form an integral part of government regulatory frameworks. However, given the constraints on budgetary resources, improving the efficiency of food safety sampling inspections (EFSSIs) remains a considerable challenge in China's food quality and safety supervision. This study aims to apply Pareto's law, starting from the examination of food sample testing items and major hazard types, to theoretically analyze methods for improving the EFSSIs. Following the theoretical analysis, the research employs provincial food sampling data from China in 2022 to empirically validate the proposed improvement strategies. RESULTS: The research findings indicate that applying Pareto's law significantly reduces the number of items that should be tested for each food subcategory, effectively lowering testing costs for each batch of food samples. Theoretically, employing Pareto's law in sampling inspections can increase the EFSSIs to 2.78 times the current observed level. Furthermore, empirical validation using food sampling data confirms that EFSSIs can be improved to 2.12 times the existing level, consistent with theoretical predictions. CONCLUSION: Implementing Pareto's law in FSSIs facilitates the detection of more unqualified food products and non‐compliant manufacturers without additional financial burden, significantly enhancing the EFSSIs. This approach provides an innovative strategy for government to bolster their food safety management efforts. © 2024 Society of Chemical Industry. [ABSTRACT FROM AUTHOR]

A proper sampling strategy is important to obtain sufficient DNA for successful identification of aged skeletal remains. The petrous bone is the highest DNA-yielding bone in the human body. Because DNA extraction from the petrous bone is very destructive, the demand for other DNA sources is significant. When investigating aged skeletal remains, teeth are usually preserved, and recent studies have shown that DNA in teeth can be best preserved in the dental cementum that surrounds the surface of the tooth root. To extract DNA from the surface of the tooth root, a nondestructive method without grinding was used. Petrous bones and teeth from 60 archaeological adult skeletons were analyzed. The DNA yield, degree of DNA degradation, and STR typing success were compared, and the results showed higher DNA yield and higher amplification success in petrous bones, despite higher degradation of petrous bones' DNA. The greater success of petrous bones is associated with poorly preserved DNA in a quarter of the teeth analyzed. When teeth with badly preserved DNA were excluded from the statistical analysis, no differences in the success of STR loci amplification were observed even if DNA yield was higher in petrous bones, which can be explained by greater degradation of petrous bones' DNA. When teeth are well preserved, they can be used for genetically analyzing aged skeletal remains instead of petrous bones, and a rapid nondestructive extraction method can be applied to shorten the identification process and to physically preserve the biological specimen. [ABSTRACT FROM AUTHOR]

Diffusion probabilistic models have demonstrated great potential in producing realistic-looking super-resolution (SR) images. However, the realism doesn't necessarily guarantee that the SR images are faithful to the ground truth high- resolution images. This paper develops a novel training-free framework namely Iterative Refinement with Controllable Condition (CCIR), for face SR based on controllable prior conditions in diffusion model. The goal is to generate SR images that are both realistic and faithful to the ground truth by controlling the prior conditions. Our framework consists of a pre-trained SR network, Local Implicit Image Function (LIIF), and a pre-trained diffusion model. The LIIF enhances the conditions provided by low-resolution images, while the diffusion model recovers fine details in the SR images. Notably, for the diffusion model, we propose a non-uniform low-pass filtering sampling strategy that dynamically adds controllable conditions to latent features during sampling process. This strategy provides a flexible balance between fidelity and realism in SR images, enabling the restoration of highly similar SR images from the same low-resolution input with different noise samples. Extensive experiments conducted on the benchmark of facial SR task demonstrate CCIR outperforms the state-of-the-art SISR methods, in qualitative and quantitative assessments, particularly in the case of magnifying very-low-resolution images or high-magnification factors. [ABSTRACT FROM AUTHOR]

Accurate recognition and localization of 3D objects is a fundamental research problem in 3D computer vision. Benefiting from transformation-free point cloud processing and flexible receptive fields, point-based methods have become accurate in 3D point cloud modeling, but still fall behind voxel-based competitors in 3D detection. We observe that the set abstraction module, commonly utilized by point-based methods for downsampling points, tends to retain excessive irrelevant background information, thus hindering the effective learning of features for object detection tasks. To address this issue, we propose MSSA, a Multi-representation Semantics-augmented Set Abstraction for 3D object detection. Specifically, we first design a backbone network to encode different representation features of point clouds, which extracts point-wise features through PointNet to preserve fine-grained geometric structure features, and adopts VoxelNet to extract voxel features and BEV features to enhance the semantic features of key points. Second, to efficiently fuse different representation features of keypoints, we propose a Point feature-guided Voxel feature and BEV feature fusion (PVB-Fusion) module to adaptively fuse multi-representation features and remove noise. At last, a novel Multi-representation Semantic-guided Farthest Point Sampling (MS-FPS) algorithm is designed to help set abstraction modules progressively downsample point clouds, thereby improving instance recall and detection performance with more important foreground points. We evaluate MSSA on the widely used KITTI dataset and the more challenging nuScenes dataset. Experimental results show that compared to PointRCNN, our method improves the AP of "moderate" level for three classes of objects by 7.02%, 6.76%, and 5.44%, respectively. Compared to the advanced point-voxel-based method PV-RCNN, our method improves the AP of "moderate" level by 1.23%, 2.84%, and 0.55% for the three classes, respectively. [ABSTRACT FROM AUTHOR]

By using 14 SSR primer pairs, we here analyzed and compared the amplification results of 534 DNA samples from six red sugar beet germplasm resources under three treatments. These data were used to explore the sampling strategy for the aforementioned resources. Based on the sampling strategy results, 21 SSR primer pairs were used to evaluate the genetic diversity of 47 red sugar beet germplasm resources. The six population genetic parameters used for exploring the sampling strategy unveiled that individual plants within the population had a relatively large genetic distance. The genetic parameters Ne, I, and Nei's of the randomly mixed sampling samples increased rapidly between 10 and 30 plants before decreasing. Therefore, when SSR technology was used to analyze the genetic diversity of the red sugar beet germplasm resources, the optimal sampling gradient for each population was the adoption of a random single-plant mixed sampling sample of no less than 10 plants and no more than 30 plants. The 21 SSR primer pairs were used to detect genetic diversity in 30 random mixed samples of 47 resources. The average polymorphic information content (PIC) was 0.5738, the average number of observed alleles (Na) was 4.1905, the average number of effective alleles (Ne) was 2.8962, the average Shannon's information index (I) was 1.1299, the average expected heterozygosity (Nei's) was 0.6127, and the average expected heterozygosity (He) was 0.6127. The genetic distance of the 47 germplasm resources ranged from 0.0225 to 0.551 (average: 0.316). According to the population structure analysis, the most suitable K value was six, which indicated the presence of six populations. Based on the clustering analysis results, the 47 germplasm resources were segregated into six groups, with obvious clustering and some germplasm resources noted for having groups with close genetic relationships. We here established a more accurate and scientific sampling strategy for analyzing the genetic diversity of red sugar beet germplasm resources by using SSR molecular markers. The findings provide a reference for collecting and preserving red sugar beet germplasms and protecting their genetic diversity. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Computer-Aided Design & Computer Graphics / Jisuanji Fuzhu Sheji Yu Tuxingxue Xuebao is the property of Gai Kan Bian Wei Hui 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.)

This paper focuses on large-scale landslide susceptibility modelling in NW Croatia. The objective of this research was to provide new insight into stable and unstable area sampling strategies on a representative inventory of small and shallow landslides mainly occurring in soil and soft rock. Four strategies were tested for stable area sampling (random points, stable area polygon, stable polygon buffering and stable area centroid) in combination with four strategies for unstable area sampling (landslide polygon, smoothing digital terrain model derived landslide conditioning factors, polygon buffering and landslide centroid), resulting in eight sampling scenarios. Using Logistic Regression, Neural Network, Random Forest and Support Vector Machine algorithm, 32 models were derived and analysed. The main conclusions reveal that polygon sampling of unstable areas is an imperative in large-scale modelling, as well as that subjective and/or biased stable area sampling leads to misleading models. Moreover, Random Forest and Neural Network proved to be more favourable methods (0.804 and 0.805 AUC, respectively), but also showed extreme sensitivity to the tested sampling strategies. In the comprehensive comparison, the advantages and disadvantages of 32 derived models were analysed through quantitative and qualitative parameters to highlight their application to large-scale landslide zonation. The results yielded by this research are beneficial to the susceptibility modelling step in large-scale landslide susceptibility assessments as they enable the derivation of more reliable zonation maps applicable to spatial and urban planning systems. [ABSTRACT FROM AUTHOR]

Most wild rose species in the Netherlands belong to Rosa section Caninae (dogroses), with Rosa arvensis (section Synstylae) and Rosa spinosissima (section Pimpinellifoliae) as other indigenous species. All species are rare, often found in small populations or as scattered individuals, except for Rosa canina and Rosa corymbifera. Conservation strategies have been developed for these roses, with a focus on ex situ methods, including clonal archives and seed orchards, using vegetative propagation from the original shrubs. Efficient collection management aims at preservation of maximum genetic diversity with a minimum of duplicated genotypes. However, dogrose taxonomy is complex because of species hybridization, different ploidy levels, and their matroclinal inheritance due to Canina meiosis. They can also reproduce vegetatively through root suckers. In order to assess the genetic structure and the levels of genetic diversity and clonality within and among the wild rose populations in the Netherlands, we genotyped individuals in wild populations and accessions in the ex situ gene bank with 10 highly polymorphic microsatellite markers. The analysis revealed 337 distinct multilocus genotypes (MLGs) from 511 sampled individuals, with some MLGs shared across different species and sites. The genetic structure analysis showed distinct clusters separating non-dogrose species from the Caninae section. Geographic distribution of MLGs indicated both local and widespread occurrences. Redundancy analysis identified 152 distinct MLGs from 244 gene bank accessions, suggesting a 38% redundancy rate. Core collections were optimized to retain genetic diversity with minimal redundancy, selecting subsets of 20–40 individuals from different species groups. The study highlights the value of genetic characterization in guiding sampling strategies for dogroses. We propose a two-step approach that may be used to reveal clonality and redundancy and to optimize core collections of species that combine sexual and vegetative reproduction, to maximize genetic capture in ex situ gene banks. [ABSTRACT FROM AUTHOR]

Copyright of Methods in Ecology & Evolution is the property of Wiley-Blackwell 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.)

Objective: Efficiently managing sea lice on salmon farms through active surveillance, crucial for lice abundance estimation, is challenging due to the need for effective sampling schemes. To address this, we developed an application that considers infestation levels, farm structure, and management protocols, enhancing the precision of sampling strategies for sea louse abundance estimation. Methods: Simulation‐based methods are valuable for estimating suitable sample sizes in complex studies where standard formulae are inadequate. We introduce FishSampling, an open Web‐based application tailored to determine precise sample sizes for specific scenarios and objectives. Result: The model incorporates factors such as sea lice abundance, farm pen numbers, potential clustering effects among these pens, and the desired confidence level. Simulation outcomes within this application provide practical advice on how to decide on the number of fish and pens to sample, under varying levels of assumed clustering. Conclusion: This approach can be used across the salmon aquaculture sector to improve sampling strategies for sea lice abundance estimation and balance surveillance costs against health objectives. Impact statementThe open‐source application FishSampling enhances sea lice monitoring on salmon farms with a novel simulation‐based approach for sample size determination. It offers precise estimates of sea lice abundance, crucial for regulatory purposes, and aids in the efficient allocation of sampling resources. [ABSTRACT FROM AUTHOR]

Surrogate-based aerodynamic robust design optimization uses a surrogate model to calculate the robustness indices, which strongly relies on the overall accuracy of the model, as well as the efficient exploration of the design space. For most surrogate modeling approaches, significant inaccuracies are often observed at the outlier region of the design space, where very few samples are spotted. A novel method using Chebyshev transformation is applied to re-allocate the orthogonal Latin hypercube sample set to alleviate the corner errors and eventually improve the overall accuracy. An inner Kriging model is developed using the sampling method, and robustness indices are calculated based on the subspaces adjacent to the sampling points. Subsequently, an outer robust model is constructed with the robustness indices as the target. Ultimately, a combination of the inner and outer models is utilized with the genetic algorithm to accomplish multi-objective robust optimization. Theoretical tests are undertaken for classic test functions, showing the advantage of the proposed approach. Based on this method, aerodynamic robust design optimizations are carried out on the RAE 2822 airfoil, for which the lift coefficient and drag coefficient are optimized for a given range of geometrical parameters. An increase of 1.94% lift coefficient and a reduction of 2.53% drag coefficient are achieved compared to the baseline design without sacrificing the robust performances. [ABSTRACT FROM AUTHOR]

Surrogate model assisted evolutionary algorithms (SAEAs) are strategies widely applied to deal with computationally expensive optimization problems (CEOPs). These methods employ metamodels to drive an evolutionary algorithm (EA) to promising design regions where new evaluations on the true-objective function must be performed. To do this, SAEAs are required to handle the challenge of training a metamodel to improve its predictions. The reliability of a metamodel is strongly related to the samples used for its training. Despite this, several SAEAs are proposed without concern about the sampling strategy employed. The ideal situation is to obtain a sample not far away from the solutions predicted on the metamodel. In this sense, the contribution/novelty of this paper regards an investigative study to compare five strategies for defining the metamodel sample in a proposed SAEA Framework (SAEA/F). The SAEA/F uses a one-dimensional Ordinary Kriging (OK) metamodel, and an expected improvement (EI) merit function is applied to define on which solutions to spend the budget of true-function evaluation. In this investigation, each strategy is incorporated into SAEA/F and then used to solve a set of analytical functions of single-objective optimization problems. The computational results suggest that two of the five sampling strategies stand out the best. The first strategy chooses those solutions with the lowest distance to the centroid of the population of solutions, and the second selects the newest solutions evaluated on true-objective function. The results highlight the potential of these approaches for solving expensive optimization problems since they speed up the algorithm convergence to improved solutions. [ABSTRACT FROM AUTHOR]

The probabilistic evaluation of the capacity and response of structural systems is often done by probabilistic simulation. This approach relies on a considerable number of deterministic analyses to obtain an accurate estimation of the desired statistical characteristics. However, the need of performing a large number of analyses hinders the application of probabilistic simulation in practice, especially when the deterministic calculation of the structural response is computationally expensive in itself (e.g. nonlinear finite element analysis). To increase the appeal of probabilistic simulation, the number of analyses should be very small (i.e. ten or less). This study shows how different sampling techniques can be adopted to select a very small sample subset of analyses to be run, and how they can yield accurate results in estimating the shear capacity of prestressed reinforced concrete beams. Based on the results of the study, the Fractile Based Sampling (FBS) method emerges as a more promising sampling strategy than other sampling techniques, like Latin Hypercube Sampling (LHS). For the same small number of samples, the results show that FBS is preferable over LHS and other sampling techniques for capturing both the general distribution of the response and the tails of the distribution, which shows its usefulness in assessing probability of failure and reliability. [ABSTRACT FROM AUTHOR]

Introduction: Native bee communities of temperate forests are conventionally sampled from the understory, yet there is growing evidence that bee assemblages in forest canopies are distinct from those in the understory. Therefore, conventional approaches to quantify forest bee-habitat relationships may not comprehensively characterize forest bee communities. Methods: To examine this, we sampled bees 1--26 m from ground level at 5-m increments at 47 locations in forests located in western Massachusetts, USA. We evaluated bee abundance and species richness responses to a suite of environmental factors measured in the understory with linear and segmented regression comparing four bee sampling strategies: (1) understory sampling only, (2) understory and midstory, (3) understory and canopy, and (4) all strata combined. Results: We found that not sampling higher strata underestimated bee abundance and species richness, and linear models had less ability to explain the data when bees of higher strata were included. Among strategies, responses analyzed linearly differed in magnitude due to overall differences in abundance and species richness, but segmented regressions showed relationships with understory characteristics that also differed in slope, which would alter interpretation. Discussion: Collectively, our findings highlight the value of including vertically stratified sampling strategies throughout the flight season to fully characterize native bee and other pollinator communities of forests. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Shenyang Agricultural University is the property of Journal of Shenyang Agricultural University Editorial Department 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.)

This article proposes a performance measure to evaluate the detection performance of a control chart with a given sampling strategy for finite or small samples sequence and prove that the CUSUM control chart with dynamic non-random control limit and a given sampling strategy can be optimal under the measure. Numerical simulations and real data for an earthquake are provided to illustrate that for different sampling strategies, the CUSUM chart will have different monitoring performance in change-point detection. Among the six sampling strategies that take only a part of samples, the numerical comparing results illustrate that the uniform sampling strategy (uniformly dispersed sampling strategy) has the best monitoring effect. [ABSTRACT FROM AUTHOR]

As part of the energy design synthesis method, complex dynamic building simulation database was created with IDA ICE code for all family house building configurations for a considered problem. In this paper, the annual heat energy demand output parameter is considered to serve as basis of a building energy design investigation. The sensitivity analysis performed by Morris' elementary effect method was used. As the result of the sensitivity analysis of the output parameter, the most important input parameters can be identified, that influence the buildings' energy efficiency, that can support further building designs. [ABSTRACT FROM AUTHOR]

During the first phase of the COVID‐19 pandemic, Istat performed the quick survey "Situation and perspectives of Italian enterprises during the COVID‐19 health emergency," with the aim of assessing the economic situation and the specific actions adopted by businesses to reduce the economic impacts of the emergency. To ensure the continuity in the information flow and to analyze the temporal evolution of the observed phenomena, the survey has been repeated in three different waves. The outcomes of each wave was released just after 2 months from the launch of the survey. The present work analyses the characteristics of the sampling strategy and describes the complexity of the data editing process, in the case of a survey planned to produce estimates able to ensure an acceptable level of accuracy in the maximum timeliness. [ABSTRACT FROM AUTHOR]

The Jbel Ressas site is an ancient Pb-Zn mine located in the northwest of Tunisia. The long activity of the mine has produced a huge volume of waste (3 tailings) rich in bioavailable fractions of Pb+Zn+Cd which have heavily polluted the surrounding agricultural soil. However, the results obtained from previous studies of this site are so variable that some authors have considered it unpolluted. It was suspected that the improper choice of sampling strategy and the extraction procedures for evaluation of the labile toxic trace element fractions were the sources of these differences. Therefore, the objective of this comparative study was to prove the interest of the choice of the sampling strategy and the extraction procedure in having a representative evaluation of mobile pollutants fractions. The critical step in any geochemical study is having representative site samples and considering that each site is unique. Therefore, environmental concerns, soil texture, principal wind directions, land slopes and topography, run flow directions, and equidistance's between the sampled plots were considered. Based on the mean total Pb+Zn+Cd concentrations obtained from analyzed samples collected in 2009, 2017, and 2021 (the present study) and previous results, the regularity and irregularity over time of the chronological evolution of the physiochemical parameters and the pollution status, respectively, were strong arguments that proved the right choice of our sampling strategy. To select the appropriate extraction procedure for labile TTE fractions, the commonly used procedures, including those used in previous studies, were tested. Using simple mode extractions, the previous authors extracted variable and even very small quantities of labile TTE, compared to similar contexts, and considered that the polluted Jbel Ressas site presented no or a low environmental risk. However, the BCR (European Community Bureau of Reference) procedure used in the present studies leached high quantities of the available toxic trace elements, which confirmed the high pollution level of the site and its environmental threat. The significant correlations among soil physicochemical parameters and total and labile TTE contents of the present study, deduced from statistical analysis, confirmed that the selected sampling strategy and the selected extraction procedure provided representative samples of the site and a fairly accurate assessment of its pollution levels. [ABSTRACT FROM AUTHOR]

The conservation of genetic diversity is a crucial aspect of forest tree breeding programs, necessitating strategies for its safeguard. Here, the extent of genetic diversity was assessed in 260 Chinese pine (Pinus tabuliformis Carr.) germplasm samples from five provenances using 24 SSR markers. We systematically compared various methods for constructing a core collection aimed at conserving genetic diversity and the results revealed substantial genetic diversity within this germplasm collection. Extensive gene exchange was observed among four of the sampled five provenances which resulted in forming two genetically distinctive groups. To construct the core collection, six different sampling strategies (PowerCore, Power marker_allele number, Power marker_gene entropy, Power marker_gene diversity, Corehunter, and genetic distance-based) and five different sampling sizes (ranging from 10 to 30%) were employed. Comparative analysis of genetic diversity parameters was conducted across the identified 26 subsets, utilizing the PowerCore strategy as the primary approach for capturing all allelic variation present in the core collection, which consisted of only 61 individuals. A supplementary collection of 20 individuals with high genetic variation was identified to provide a final core collection of 81 individuals, representing 31.2% of the initial collection. The constructed core collection effectively captured the genetic diversity present in the initial collection and serves as a valuable resource for preserving genetic richness within the breeding population. [ABSTRACT FROM AUTHOR]

The development of high-throughput sequencing (HTS) of environmental DNA (eDNA) has stimulated the study of soil microbial diversity patterns and drivers at all scales. However, given the heterogeneity of soils, a challenge is to define effective and efficient sampling protocols that allow sound comparison with other records, especially vegetation. In studies of elevational diversity pattern, a trade-off is choosing between replication within elevation bands vs. sampling more elevation bands. We addressed this question for soil protists along an elevation gradient on Mt. Asahi, Hokkaido, Japan. We compared two sampling approaches: (1) the replicate strategy (five replicates at six elevational bands, total = 30) and (2) the transect strategy (one sample in each of 16 different elevational bands). Despite a nearly twofold lower sampling effort, the transect strategy yielded congruent results compared to the replicate strategy for the estimation of elevational alpha diversity pattern: the regression coefficients between diversity indices and elevation did not differ between the two options. Furthermore, for a given total number of samples, gamma diversity estimated across the entire transect was higher when sampling more elevational bands as compared to replication from fewer elevational bands. Beta diversity (community composition turnover) was lower within a given elevational band than between adjacent bands and increased with elevation distance. In redundancy analyses, soil organic matter-related variable (the first principal component of soil organic matter, water content, total organic carbon, and nitrogen by whom were highly correlated) and elevation best explained elevational beta diversity pattern for both sampling approaches. Taken together, our results suggest that sampling a single plot per elevation band will be sufficient to obtain a good estimate of soil micro-eukaryotic diversity patterns along elevation gradients. This study demonstrated the effectiveness of the transect strategy in estimating diversity patterns along elevation gradients which is instructive for future environmental or even experimental studies. While not advocating for completely replacing replication-based sampling practices, it is important to note that both replicate and transect strategies have their merits and can be employed based on specific research goals and resource limitations. [ABSTRACT FROM AUTHOR]

A Reconfigurable Modular Robotic System (RMRS) consists of multiple interconnected robots and can achieve various functionalities by rearranging its modular robots, such as transporting loads of various shapes. The path planning for an RMRS involves the system motion and also its formation arrangements. Sampling-based path planning for the RMRS might be inefficient due to the formation variety. Recently, convex subsets of the obstacle-free workspace, referred to as polygon nodes, are instead sampled to formulate constrained optimization problems. The success rate of sampling is however unsatisfactory due to connectivity requirements. This paper proposes an obstacle-aware mixture density network to guide the generation of polygon nodes, where the connectivity of polygon nodes is guaranteed by non-zero Minkowski differences between the formation geometry and the intersection of nodes. Subsequently, Convex-Polygon Trees* (CPTs*) are proposed to connect these polygon nodes in an RRT* manner, outputting candidates of convex optimization problems. The optimality degeneration due to distance approximation is proven bounded and the computational complexity is shown linear to the Lebesgue measure of the entire workspace space. Numerical simulations have shown that in most tested large and cluttered environments the CPT* is more than 8 times faster than an existing constrained optimization method. The results have also shown CPT*’ improved scalability to large environments and enhanced efficiency in dealing with narrow passages. [ABSTRACT FROM AUTHOR]

Pitfall traps are frequently used to capture ground‐dwelling arthropods, particularly beetles, ants and spiders. The capture efficiency of a pitfall trapping system strongly depends on the number and opening size of traps, how traps are distributed over the sampling area (spatial arrangement) and the movement characteristics of arthropods.We use numerical simulations for a single species to analyse the trap count patterns that emerge from these variables. Arthropod movement of individuals is modelled as correlated random walks, with multiple traps placed over an area, and catches are simulated as individual interaction with traps. We consider four different types of spatial arrangements of traps across a homogeneous landscape: grid (i.e. rectangular array), transect, nested‐cross and randomised. We contextualise our results by considering the locomotion of Pterostichus melanarius, a highly active carabid beetle often serving as a biocontrol agent for the suppression of pest insects and weeds.By simulating the trapping of randomly moving ground‐dwelling arthropods, we show that there is an optimal inter‐trap separation distance (trap spacing) that maximises captures, that can be expressed using exact formulae in terms of trap opening sizes, sampling area and trap number. Moreover, for the grid and nested‐cross arrangements, larger trap spacing to maximise spatial coverage over the whole sampling area is suboptimal. Also, we find that over a large sampling area, there is a hierarchical order for spatial arrangements in relation to capture efficiency: grid, randomised, transect, followed by the nested‐cross. However, over smaller sampling areas, this order is changed as the rate at which trap counts accumulate with trap number varies across arrangements—eventually saturating at different levels. In terms of movement effects, capture efficiency is maximised over a narrow diffusive range and does not depend strongly on the type of spatial arrangement—indicating an approximate optimal mode of arthropod activity, i.e. rate of spread.Our approach simultaneously considers several important experimental design aspects of pitfall trapping providing a basis to optimise and adapt sampling protocols to other types of traps to better reflect their various purposes, such as monitoring, conservation or pest management. [ABSTRACT FROM AUTHOR]

The determination of sample size in qualitative research introduces a unique and multifaceted challenge, setting it apart from the more structured methodology of quantitative research. Contrary to sampling methods in quantitative research, which primarily aim to secure random and statistically representative samples that facilitate the generalisation of findings to broader populations, sampling in qualitative research requires a distinct set of considerations in its pursuit of a deeper understanding of specific phenomena. The objective of this editorial is to provide qualitative researchers with clear and foundational guidance for effectively communicating the methodological aspects of their research papers, particularly pertaining to sample size justification. Building on this, we present S.C.A.D.E, an acronym comprising five key actionable elements--Selecting, Clarifying, Aligning, Deploying and Evaluating--to guide researchers in determining the appropriate sample size and ensuring that data saturation is achieved as they plan their qualitative exploration. [ABSTRACT FROM AUTHOR]

Since the implementation of new EU limits on cadmium (Cd) in cacao-derived products, reliable measurements of the Cd concentration in cacao samples have become even more important. This study was set up to analyse the robustness of the measured Cd concentrations in cacao as affected by sampling strategy and by the laboratory receiving these samples. Six different homogenised cacao liquor samples were sent to 25 laboratories, mainly located in Latin America. On average, only 76% of the laboratories reported acceptable results per sample using internationally accepted criteria. More unreliable data was obtained when Atomic Absorption Spectroscopy (AAS) rather than Inductively Coupled Plasma (ICP) instruments were used or where concentrations were outside the calibration range. Subsequently, four commercial lots in Ecuadorian warehouses were sampled to identify the variation among beans, bags and replicate chemical analyses of ground samples. Simulations indicate that a composite sample should be made from at least 10 bags on a pallet and at least 60 beans should be ground prior to analysis to obtain an acceptable CV below 15%. This study shows that current Cd analyses in cacao on the market are neither sufficiently accurate nor precise and that more control on laboratory certifications is needed for reliable screening of Cd in cacao. [ABSTRACT FROM AUTHOR]

Regional modeling of landslide hazards is an essential tool for the assessment and management of risk in mountain environments. Previous studies that have focused on modeling earthquake-triggered landslides report high prediction accuracies. However, it is common to use a validation strategy with an equal number of landslide and non-landslide samples, scattered homogeneously across the study area. Consequently, there are overestimations in the epicenter area, and the spatial pattern of modeled locations does not agree well with real events. In order to improve landslide hazard mapping, we proposed a spatially heterogeneous non-landslide sampling strategy by considering local ratios of landslide to non-landslide area. Coseismic landslides triggered by the 2008 Wenchuan Earthquake on the eastern Tibetan Plateau were used as an example. To assess the performance of the new strategy, we trained two random forest models that shared the same hyperparameters. The first was trained using samples from the new heterogeneous strategy, and the second used the traditional approach. In each case the spatial match between modeled and measured (interpreted) landslides was examined by scatterplot, with a 2 km-by-2 km fishnet. Although the traditional approach achieved higher AUCROC (0.95) accuracy than the proposed one (0.85), the coefficient of determination (R2) for the new strategy (0.88) was much higher than for the traditional strategy (0.55). Our results indicate that the proposed strategy outperforms the traditional one when comparing against landslide inventory data. Our work demonstrates that higher prediction accuracies in landslide hazard modeling may be deceptive, and validation of the modeled spatial pattern should be prioritized. The proposed method may also be used to improve the mapping of precipitation-induced landslides. Application of the proposed strategy could benefit precise assessment of landslide risks in mountain environments. [ABSTRACT FROM AUTHOR]

Image registration is the basis for the joint interpretation of synthetic aperture radar (SAR) and optical images. However, the significant nonlinear radiation difference (NRD) and the geometric imaging model difference render the registration quite challenging. To solve this problem, both traditional and deep learning methods are used to extract structural information with dense descriptions of the images, but they ignore that structural information of the image pair is coupled and often process images separately. In this paper, a deep learning-based registration method with a co-attention matching module (CAMM) for SAR and optical images is proposed, which integrates structural feature maps of the image pair to extract keypoints of a single image. First, joint feature detection and description are carried out densely in both images, for which the features are robust to radiation and geometric variation. Then, a CAMM is used to integrate both images' structural features and generate the final keypoint feature maps so that the extracted keypoints are more distinctive and repeatable, which is beneficial to global registration. Finally, considering the difference in the imaging mechanism between SAR and optical images, this paper proposes a new sampling strategy that selects positive samples from the ground-truth position's neighborhood and augments negative samples by randomly sampling distractors in the corresponding image, which makes positive samples more accurate and negative samples more abundant. The experimental results show that the proposed method can significantly improve the accuracy of SAR–optical image registration. Compared to the existing conventional and deep learning methods, the proposed method yields a detector with better repeatability and a descriptor with stronger modality-invariant feature representation. [ABSTRACT FROM AUTHOR]

In deep learning-based hyperspectral remote sensing image classification tasks, random sampling strategies are typically used to train model parameters for testing and evaluation. However, this approach leads to strong spatial autocorrelation between the training set samples and the surrounding test set samples, and some unlabeled test set data directly participate in the training of the network. This leaked information makes the model overly optimistic. Models trained under these conditions tend to overfit to a single dataset, which limits the range of practical applications. This paper analyzes the causes and effects of information leakage and summarizes the methods from existing models to mitigate the effects of information leakage. Specifically, this paper states the main issues in this area, where the issue of information leakage is addressed in detail. Second, some algorithms and related models used to mitigate information leakage are categorized, including reducing the number of training samples, using spatially disjoint sampling strategies, few-shot learning, and unsupervised learning. These models and methods are classified according to the sample-related phase and the feature extraction phase. Finally, several representative hyperspectral image classification models experiments are conducted on the common datasets and their effectiveness in mitigating information leakage is analyzed. [ABSTRACT FROM AUTHOR]

Contaminated sites pose a serious threat to the ecological environment and human health. Because of the presence of multiple peaks in the pollution data of some contaminated sites, as well as strong spatial heterogeneity and skewness in their distribution, the accuracy of spatial interpolation prediction is low. This study proposes a method for investigating highly skewed contaminated sites, which uses Thiessen polygons coupled with geostatistics and deterministic interpolation to optimize the spatial prediction and sampling strategy of sites. An industrial site in Luohe is used as an example to validate the proposed method. The results indicate that using 40 × 40 m as the minimum initial sampling unit can obtain data that is representative of the regional pollution situation. Evaluation indexes reveal that the ordinary kriging (OK) method for interpolation prediction accuracy and the radial basis function_inverse distance weighted (RBF_IMQ) method for pollution scope prediction provides the best results, which can effectively improve the spatial prediction accuracy of pollution in the study area. Each accuracy indicator is enhanced by 20–70% after supplementing 11 sampling points in the suspect region, and the identification of the pollution scope approaches 95%. This method offers a novel approach for investigating highly biased contaminated sites, which can optimize the spatial prediction accuracy of pollution and reduce economic costs. [ABSTRACT FROM AUTHOR]

This study aims to establish an accurate prediction model using artificial neural networks (ANNs) to effectively and efficiently predict the process-induced warpage of a flip-chip chip-scale package (FCCSP). To enhance model performance, a novel subdomain-based sampling strategy and Taguchi hyperparameter optimization are proposed in the ANN algorithm. To simulate the warpage behavior the FCCSP during fabrication, a process modeling approach is proposed, where the viscoelastic behavior of the epoxy molding compound is included, in which the viscoelastic properties are determined using dynamic mechanical measurement. In addition, the temperature-dependent thermal-mechanical properties of the materials in the FCCSP are assessed through thermal-mechanical analysis and dynamic mechanical analysis. The modeled warpage results are verified by the warpage measurement. Next, warpage parametric analysis is performed to identify the key factors most affecting warpage behavior for use in the construction of the warpage prediction model. Moreover, the advantages of the proposed sampling and hyperparameter tuning approaches are proved by comparing with other existing models, and the validity of the developed ANN-based deep learning warpage prediction model is demonstrated through a validation dataset. [ABSTRACT FROM AUTHOR]

In this paper, we are concerned with microwave subsurface imaging achieved by inverting the linearized scattering operator arising from the Born approximation. In particular, we consider the important question of reducing the required data to achieve imaging. This can help to reduce the radar system's cost and complexity and mitigate the imaging algorithm's computational burden and the needed storage resources. To cope with these issues, in the framework of a multi-monostatic/multi-frequency configuration, we introduce a new spatial sampling scheme, named the warping method, that allows for a significant reduction in spatial measurements compared to other literature approaches. The basic idea is to introduce some variable transformations that "warp" the measurement space so that the reconstruction point-spread function obtained by adjoint inversion is recast as a Fourier-like transformation, which provides insights into how to achieve the sampling. In our previous contributions, we focused on presenting and checking the theoretical background with simple numerical examples. In this contribution, we briefly review the key components of the warping method and present its experimental validation by considering a realistic subsurface scattering scenario for the case of a buried water pipe. Essentially, we show that the latter succeeds in reducing the number of data compared to other approaches in the literature, without significantly affecting the reconstruction results. [ABSTRACT FROM AUTHOR]

In the realm of urban transportation management, identifying critical locations within transportation networks is paramount for efficient urban planning, crisis management, and infrastructure development. This study investigates the effectiveness of the Network Kernel Density Estimation (Net-KDE) method in identifying these critical locations. The Net-KDE method is chosen for its inherent capacity to consider spatial patterns and path length between sample and estimation points, making it well-suited for capturing the complexities of urban transportation networks. The approach is experimented with a series of control maps applied to Tehran city. Overall, Net-KDE provides valuable and agreeable outputs that show its potential. The findings underscore the crucial role of scenario customization in enhancing the method's accuracy. Remarkably, featuring adaptive bandwidth and strategic sampling emerges as the most effective in identifying critical locations. This research has a broader impact on crisis responders, urban developers, and city planners in addition to improving our knowledge of urban network dynamics. [ABSTRACT FROM AUTHOR]

Unsupervised domain adaptation has gotten a lot of attention due to its ability to improve learning performance in a target domain based on the knowledge extracted from a source domain. Recent studies show that graph-based models can accomplish good results for domain adaptation problems. However, most of these graph-based domain adaptation approaches cannot work in an end-to-end manner, leading to the limited scalable. To address this issue, we propose a learning method named Mini-batch Dynamic Geometric Embedding (MDGE), which seeks to find the relationship between batches source and target samples to learn discriminative representations. Specifically, to build a better graph representing sample relationship, we propose a class-specific sampling strategy to pick up samples which are then used as input of MDGE. Since the samples are effectively selected, we develop a method to dynamically build a subgraph that in turn supports the relationship update and helps the network backbone to extract more discriminative features. Comprehensive experiments on real-world visual datasets demonstrate the effectiveness of the proposed MDGE algorithm. [ABSTRACT FROM AUTHOR]

Due to the low number of bacteraemias identified after the first day of fever, repeating blood cultures adds little additional information while potentially increasing iatrogenesis, costs, and the number of false positive blood cultures. Repeat blood cultures in children with persistent fever and neutropenia: diagnostic and clinical implications: repeat blood cultures for fever and neutropenia. The new blood culture sample strategy - four adequately filled blood culture bottles on day 1 - efficiently improved the diagnosis of bacteriemia over the three periods. [Extracted from the article]

Toona sinensis is an important "vegetable and timber" tree species that is widely distributed in eastern and southeastern Asia. An efficient mechanism for preserving and utilizing germplasm resources is crucial for speeding up the process of genetic improvement of T. sinensis. We collected a total of 1040 germplasm resources of T. sinensis from China, Britain, and New Zealand. Using a total of 27,040 markers, we analyzed the genetic diversity, genetic structure, and core collection of these resources. Additionally, a comprehensive evaluation method for phenotypes of T. sinensis was established by principal component analysis (PCA) and the technique for order preference by similarity to ideal solution (TOPSIS). Our analysis included three main results: Firstly, the breeding population, consisting of a total of 1040 individuals, contains a diverse genetic pool and can be divided into two genetic clusters. Secondly, the best core collection was obtained from the alternative population, consisting of a total of 208 individuals and comprising 20% of the breeding population. It was screened out based on Core Hunter 3 software with the average entry-to-nearest-entry (E-NE) algorithm. Finally, a comprehensive evaluation method was established based on eight selected agronomic traits of the core collection. This method allowed us to screen the best individual plants of the core collection, including N0237, N0972, and N0409. Overall, our research findings will enable better preservation and utilization of T. sinensis germplasm resources and result in faster improvement of T. sinensis varieties. [ABSTRACT FROM AUTHOR]

Background Surveys for intestinal parasitic infections (IPIs) often involve samples from three sampling dates analysed by various microscopy techniques. However, analysis of three samples per individual is a huge burden on time and resources. We compared the value from analysing three or fewer samples. Methods In this cross-sectional study, three faecal samples were collected every other day from 332 children from two locations in Guantanamo province, Cuba. Samples were analysed by wet mount with Lugol stain, Willis flotation method and Kato–Katz thick smear. Results Most parasites were detected by wet mount, although helminth eggs not found by wet smear were detected by the Willis flotation method (in particular) and Kato–Katz smear. Blastocystis spp. was the most commonly detected parasite (about 65%), then Giardia duodenalis and then Entamoeba spp. Although analysis of two stool samples significantly increased occurrence data for Blastocystis , this was not so for the other parasites. For none of the protozoan parasites were results from analysing three samples significantly higher than results from analysing just two samples. Conclusions Analysing two faecal samples by wet mount and the Willis flotation method provides useful data for estimating the prevalence of IPIs in relatively high prevalence settings. Analysing further samples provides limited additional information and adds an extra burden in terms of time and resources. [ABSTRACT FROM AUTHOR]

The selection of unburned labels is a crucial step in machine learning modelling of wildfire occurrence probability. However, the effect of different sampling strategies on the performance of machine learning methods has not yet been thoroughly investigated. Additionally, whether the ratio of burned labels to unburned labels should be balanced or imbalanced remains a controversial issue. To address these gaps in the literature, we examined the effects of four broadly used sampling strategies for unburned label selection: (1) random selection in the unburned areas, (2) selection of areas with only one fire event, (3) selection of barren areas, and (4) selection of areas determined by the semi-variogram geostatistical technique. The effect of the balanced and imbalanced ratio between burned and unburned labels was also investigated. The random forest (RF) method explored the relationships between historical wildfires that occurred over the period between 2001 and 2020 in Yunnan Province, China, and climate, topography, fuel and anthropogenic variables. Multiple metrics demonstrated that the random selection of the unburned labels from the unburned areas with an imbalanced dataset outperformed the other three sampling strategies. Thus, we recommend this strategy to produce the required datasets for machine learning modelling of wildfire occurrence probability. This study evaluates the effect of different sampling strategies for unburned label selection and different ratios of burned and unburned labels in the datasets on the accuracy of machine learning modelling of wildfire occurrence probability. [ABSTRACT FROM AUTHOR]

Copyright of Application Research of Computers / Jisuanji Yingyong Yanjiu is the property of Application Research of Computers Edition 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.)

The recognition of fine-grained objects is crucial for future remote sensing applications, but this task is faced with the few-shot problem due to limited labeled data. In addition, the existing few-shot learning methods do not consider the unique characteristics of remote sensing objects, i.e., the complex backgrounds and the difficulty of extracting fine-grained features, leading to suboptimal performance. In this study, we developed an improved task sampling strategy for few-shot learning that optimizes the target distribution. The proposed approach incorporates broad category information, where each sample is assigned both a broad and fine category label and converts the target task distribution into a fine-grained distribution. This ensures that the model focuses on extracting fine-grained features for the corresponding broad category. We also introduce a category generation method that ensures the same number of fine-grained categories in each task to improve the model accuracy. The experimental results demonstrate that the proposed strategy outperforms the existing object recognition methods. We believe that this strategy has the potential to be applied to fine-grained few-shot object recognition, thus contributing to the development of high-precision remote sensing applications. [ABSTRACT FROM AUTHOR]

Bamboo crown width (CW) is a reliable index for evaluating growth, yield, health and vitality of bamboo, and light capture ability and carbon fixation efficiency of bamboo forests. Based on statistical results produced from fitting the eight basic growth functions using data from 1374 Phyllostachys pubescens in Yixing, Jiangsu Province, China, this study identified the most suitable function (logistic function) to construct a two-level mixed effects (NLME) CW model with the forest block and sample plot-level effects included as random effects in the model. Four methods for selecting sample bamboos per sample plot (largest bamboo, medium-sized bamboo, smallest bamboo, and randomly selected bamboos) and eight sample sizes (1–8 selected bamboos per sample plot) were evaluated to calibrate our NLME CW model. Using diameter at breast height (DBH), height to crown base (HCB), arithmetic mean diameter at breast height (MDBH), and height (H) as predictor variables, the model produced the best fit statistics (Max R2, min RMSE, and TRE). This model was further improved by introducing random effects at two levels. The results showed a positive correlation of CW with HCB and DBH and a negative correlation with H. The smallest two bamboo poles per sample plot used to estimate the random effects of the NLME model provided a satisfactory compromise regarding measurement cost, model efficiency, and prediction accuracy. The presented NLME CW model may guide effective management and carbon estimation of bamboo forests. [ABSTRACT FROM AUTHOR]

Reliable crop type classification supports the scientific basis for food security and sustainable agricultural development. However, it still lacks a limited-samples-based crop classification method which is labor- and time-efficient. To this end, we used the Google Earth Engine (GEE) and Sentinel-1A/B SAR time series to develop eight types of crop classification strategies based on different sampling methods of central and scattered, different perspectives of object-based and pixel-based, and different classifiers of the Time-Weighted Dynamic Time Warping (TWDTW) and Random Forest (RF). We carried out 30-times classifications with different samples for each strategy to classify the crop types at the North Dakota–Minnesota border in the U.S. We then compared their classification accuracies and assessed the accuracy sensitivity to sample size. The results found that the TWDTW generally performed better than RF, especially for small-sample classification. Object-based classifications had higher accuracies than pixel-based classifications, and the object-based TWDTW had the highest accuracy. RF performed better in scattered sampling than the central sampling strategy. TWDTW performed better than RF in distinguishing soybean and dry bean with similar curves. The accuracies improved for all eight classification strategies with increasing sample size, and TWDTW was more robust, while RF was more sensitive to sample size change. RF required many more samples than TWDTW to achieve satisfactory accuracy, and it performed better than TWDTW when the sample size exceeded 50. The accuracy comparisons indicated that the TWDTW has stronger temporal and spatial generalization capabilities and has high potential applications for early, historical, and limited-samples-based crop type classification. The findings of our research are worthwhile contributions to the methodology and practices of crop type classification as well as sustainable agricultural development. [ABSTRACT FROM AUTHOR]

Exploring spatial-spectral data frequently involves classifying hyperspectral images using convolutional neural networks. Due to the high complexity of the data and the scarcity of available training samples, hyperspectral image classification presents significant difficulties. In the context of supervised classification, we find that traditional experimental designs are frequently misused in the spectral-spatial processing context, resulting in unfair or biased performance evaluation, particularly when training and testing samples are selected at random from the same dataset. Under these circumstances, the dependence caused by the overlap between training and testing samples may be artificially increased, in breach of the data independence assumption upheld by supervised learning theory. In order to prevent an unbiased classification result, we present in this paper a controlled strategy designed to minimize the overlap between the samples present in the training and the testing data sets. The proposed controlled sampling strategy ensures a more trustworthy generalization of the CNN model by minimizing the issues present in the random sampling approach, such as the inability to determine whether or not an increase in classification accuracy is due to the spatial information incorporated into a classifier or to an increase in the overlap between training and testing data sets. Experiments performed with a wavelet CNN on different HSIs, namely Indian Pines, Pavia University, and Salinas, ensure the generalization of the data under the assumption that the training and data sets are independent from one another, based on a controlled strategy. Considering the high dimension of the HSI image, as a preprocessing step, the evaluation of the proposed framework is done by PCA and FA methods. [ABSTRACT FROM AUTHOR]

The dynamic precision of the spindle directly affects the machining precision and production efficiency of CNC machine tools. Given the complexity of spindle dynamic behaviors under different working conditions, the spindle dynamic precision is difficult to model accurately using theoretical methods. In this paper, a Kriging-based response surface method is proposed for spindle dynamic precision modeling and prediction under all working conditions. Then, a novel sampling strategy is proposed to improve the robustness and efficiency. Specifically, multi-zone Latin hypercube sampling (MZ-LHS) is designed with the consideration of spindle dynamic characteristics for the initial modeling, and an active learning method that combines Kriging and Monte Carlo simulation (AK-MCS) is further improved by parallel processing for the model's global updating and refinement. The proposed method ensures that the approximation of precision features can be accurately and efficiently fitted under complex working conditions without introducing traditional theoretical rotor dynamics models. The effectiveness of the proposed method is verified by conducting the workload simulation and precision measurement test on a CNC machine tool spindle, and the dynamic precision of the tested spindle is analyzed as a case study. [ABSTRACT FROM AUTHOR]

In general, the main criteria for determining the optimal soil sampling grid are the configuration of the relief and the soil properties of a particular land parcel. Soil sampling for agrochemical testing of soils is regulated by Slovak legislation, for the vineyard the density of the sampling grid is determined as one sample per two to four hectares of land, what means six to nine samples for the selected research area of 17 ha. This density is several times lower compared to the results of most scientific papers, therefore 22 samples (1 sample / 0.77 ha) were taken to determine phosphorus and magnesium and 73 samples (1 sample / 0.23 ha) to determine the carbon and nitrogen content. Subsequently, 12 dilute networks were created from the concentrated sampling network, which entered into interpolation. By comparing the created rasters, we determined the value of spatial loss of information on the state of nutrients in the soil, which made it possible to determine the optimal density of soil sampling for the needs of precision agriculture, considering the financial complexity of laboratory analyses. We determined the most suitable 10-point grid (1 sample / 1.7 ha). The research was carried out on a model vineyard in Modrany (Komárno district). [ABSTRACT FROM AUTHOR]

Many problems arising in computational science and engineering can be described in terms of approximating a smooth function of d variables, defined over an unknown domain of interest Omega subset Rd, from sample data. Here both the underlying dimensionality of the problem (in the case d≥1) as well as the lack of domain knowledge--with\Omega potentially irregular and/or disconnected--are confounding factors for sampling-based methods. Naïve approaches for such problems often lead to wasted samples and inefficient approximation schemes. For example, uniform sampling can result in upward of 20\% wasted samples in some problems considered herein. In applications such as surrogate model construction in computational uncertainty quantification, the high cost of computing samples necessitates a more efficient sampling procedure. Over the last several years methods for computing such approximations from sample data have been studied in the case of irregular domains, and the advantages of computing sampling measures depending on an approximation space P of dim(P) = N have been shown. More specifically, such approaches confer advantages such as stability and well-conditioning, with an asymptotically optimal sample complexity scaling\scrO (N log(N)). The recently proposed adaptive sampling for general domains (ASGD) strategy is one such technique to construct these sampling measures. The main contribution of this paper is a procedure to improve upon the ASGD approach by adaptively updating the sampling measure in the case of unknown domains. We achieve this by first introducing a general domain adaptivity strategy, which computes an approximation of the function and domain of interest from the sample points. Second, we propose an adaptive sampling strategy, termed adaptive sampling for unknown domains (ASUD), which generates sampling measures over a domain that may not be known in advance, based on the ideas introduced in the ASGD approach. We then derive (weighted) least squares and augmented least squares techniques for polynomial approximation on unknown domains. We present numerical experiments demonstrating the efficacy of the adaptive sampling techniques with least squares--based polynomial approximation schemes. Our results show that the ASUD approach consistently achieves errors the same as or smaller than uniform sampling, but using fewer, and often significantly fewer, function evaluations. [ABSTRACT FROM AUTHOR]

The past landslides from landslide inventory are essential for machine learning (ML)-based landslide susceptibility assessment (LSA). They determine not only the positive samples (characterized by the past landslides) but also the negative samples (randomly generated based on past landslides) for training and validating the ML models. However, the number of past landslides is often limited because of the constraints of time, budget, and resources available, etc. In other words, the available data for establishing landslide susceptibility ML models are limited, which indicates that the accuracy and reliability of the corresponding models are insufficient. This article, therefore, proposes using a new landslide susceptibility index-based sampling strategy to enhance the positive and negative samples for model training and validation to reach an improved ML-based LSA. To realize this idea, landslide susceptibility analysis based on initial datasets compiled from landslide inventory by using three ML models, i.e., random forest (RF), gradient boosting decision tree (GBDT) and support vector machine (SVM), are first conducted to obtain the initial landslide susceptibility indices at different space locations. Then, the landslide susceptibility indices are analyzed with the proposed sampling strategy which considers directly the grid units with very high and very low landslide susceptibility indices as potential positive and negative samples, respectively, to enrich the initial dataset; and the number of these positive/negative samples is determined by a Bayesian optimization algorithm. Thereafter, the ML models are updated with the enriched datasets. Finally, to verify the effectiveness of the proposed strategy, the improved models are applied to assess the landslide susceptibility of Taojiang County, China, and the results are compared with those from initial corresponding models without updating. The results show that compared with the initial RF, GBDT and SVM models, the corresponding improved models have a better performance in accuracy, precision, recall rate, and specificity. In particular, the AUC values of the three models are increased by 11.94%, 10.72% and 0.57%, respectively. [ABSTRACT FROM AUTHOR]

Various evolutionary algorithms (EAs) have been proposed to address feature selection (FS) problems, in which a large number of fitness evaluations are needed. With the rapid growth of data scales, the fitness evaluation becomes time consuming, which makes FS problems expensive optimization problems. Surrogate-assisted EAs (SAEAs) have been widely used to solve expensive optimization problems. However, the SAEAs still face difficulties in solving expensive FS problems due to their high-dimensional discrete decision variables. To address this issue, we propose an SAEA with parallel random grouping for expensive FS problems, in which three main components consist. First, a constraint-based sampling strategy is proposed, which considers the influence of the constraint boundary and the number of selected features. Second, a high-dimensional FS problem is randomly divided into several low-dimensional subproblems. Surrogate models are then constructed in these low-dimensional decision spaces. After that, all the subproblems are optimized in parallel. The process of random grouping and parallel optimization continues until the termination condition is met. Finally, a final solution is chosen from the best solution in the historical search and the best solution in the last population using a random, distance-, or voting-based method. Experimental results show that the proposed algorithm generally outperforms traditional, ensemble, and evolutionary FS methods on 14 datasets with up to 10 000 features, especially when the required number of real fitness evaluations is limited. [ABSTRACT FROM AUTHOR]

This study reports the data of polychlorinated dibenzo-p-dioxins (PCDDs), -furans (PCDFs), and polychlorinated biphenyls dioxin-like (dl PCBs) measured in a total of 260 samples of the dairy supply chain collected over a period of 7 years (2011–2017) in the Latium region (Italy). Levels and average profiles of congeners were reported for each group of the analyzed dairy matrices, and any differences between different sampling strategies were considered (around likely pollutant sources or casual sampling). Of the samples, 95.4% resulted compliant with the regulated levels; only samples belonging to the "sheep bulk milk" matrix were found to be above either the action levels or the maximum levels (tot. 12 samples). Raw milk of the sheep species showed the highest averages (PCDD/F 0.248 and dl PCB 0.966 WHO TEQ pg/g of fat) compared to the milk of other species. The buffalo milk showed a content of dl PCB significantly lower (dl PCB: 0.371 WHO TEQ pg/g of fat) than the sheep milk (p<0.05). Dioxins were found to be superior to furans in almost all dairy products, except in the noncompliant samples where furans were higher. The OCDD was found to be the most abundant congener in almost all dairy products. This study provides a first list of reference values for background contamination of the dairy supply chain in the Latium region. These pre-existing values will be useful in all cases of environmental pollution to identify critical situations. [ABSTRACT FROM AUTHOR]

Advancements in high-speed computer technology play an ever-increasing role in analyzing various types and massive size data. However, handling big high-dimensional data sets is a challenge in terms of computational storage and capacity. Through feature selection methods, data dimensions can be reduced by eliminating the dummy variables, allowing for more extensive analysis. In this paper, data are classified based on the ratio of response into two types: balanced (almost the same ratio for each class) and partially balanced (consists of majority and minority, with virtually the same ratio for minority classes). Performance comparisons of various feature selection methods for balanced and partially balanced data are provided. This approach will help in selecting sampling strategy and feature selection methods that perform well while utilizing appropriate resources for high-dimensional data. [ABSTRACT FROM AUTHOR]

Owing to the difficulties in determining the boundaries of landslides from landslide inventory, landslide samples are often expressed by geometrical points, rather than the actual shape of landslides. However, when grid units are used for landslide susceptibility assessment (LSA), the number of landslide samples is quite limited by using geometric points to represent landslides. This situation further leads to a small number of non-landslide samples that are randomly generated based on the landslide samples. This paper proposes a frequency ratio (FR)-based sampling strategy to enhance the information of landslides to achieve an improved machine learning-based LSA. To realize this idea, the FR values of landslide conditioning factors are first calculated, and each grid unit has an attribute dataset composed of FR values. Then, the grid units with the minimum FR value no less than 1 are selected as supplementary positive samples, the grid units with the maximum FR value less than 1 are selected as supplementary negative samples. The enhanced dataset is finally composed of supplementary positive samples and historical landslide points. Thereafter, two typical ML models of the random forest (RF) and support vector machine (SVM) are employed to construct the improved LSA models based on the enhanced datasets. Finally, the landslide susceptibility of Anhua County, China, is assessed by the improved ML models, and the results are compared with the ML models trained by the traditional sampling method. The results indicate that compared with traditional RF and SVM models, the corresponding improved models show better performance in accuracy, precision, recall, and F1 value. In particular, the AUC values of the two models are increased by 9% and 8%, respectively. The proposed method provides a promising alternate for an accurate and reliable LSA. [ABSTRACT FROM AUTHOR]