Plan Sample Selection & Protection

The objective of this study was to compare the efficacy of 2 dextran/swim-up media to increase the sperm quality parameters and the maintenance of these parameters at 15 degrees C and 30 degrees C over 6 hours. Additionally, this study examined whether differences in sperm quality reflect different reproductive efficiencies following intrauterine insemination in superovulated ewes. The study involved 2 selected samples (SS) obtained by dextran/swim-up, performed either with (SS+) or without (SS-) capacitating compounds, and a control sample consisting of raw semen diluted in the same medium. The efficacies of the swim-up sperm selection procedures were similar in both media, and no significant differences were found among the evaluated parameters. Conversely, we found important differences between selected and control samples. Sperm motility, viability (as assessed by carboxifluorescein diacetate/propidium iodide [PI] staining), and mitochondrial activity (as assessed by rhodamine 123/PI) were significantly higher in the selected samples than in the control. Additionally, following incubation at 15 degrees C, the preservation of sperm quality was significantly better in the selected samples than in the control samples. After 6 hours of incubation at 15 degrees C, selected samples had a motility value of 46%, which was significantly (P < .001) higher than the value observed in control samples (27%). The percentage of viable cells observed after 6 hours of incubation at 15 degrees C was significantly (P < .0001) higher in selected samples than in the control samples. Furthermore, after 2 hours of incubation at 30 degrees C, swim-up samples had viability values that were significantly (P < .0001) higher than those of the control samples. SS+ and SS- samples did not differ significantly in spermatozoa yield, sperm quality, or survival. Differences between selected samples and controls were reflected in the fertilization rate obtained following intrauterine insemination in superovulated ewes that experienced a 52-hour interval between progestagen removal and artificial insemination. A restricted criterion for fertilization rate evaluation was established, and only the percentage of embryos recovered from the uterine horns 6 days after insemination was considered with respect to the total number of corpora lutea counted in the ovaries. The fertilization rate of SS- samples (50%) was significantly higher (P > .001) than those of the SS+ (2%) and control samples (5%).

Heckman’s (Ann Econ Soc Meas 4(5), 475–492, 1976; Econometrica 47, 153–161, 1979) sample selection model has been employed in three decades of applications of linear regression studies. This paper builds on this framework to obtain a sample selection correction for the stochastic frontier model. We first show a surprisingly simple way to estimate the familiar normal-half normal stochastic frontier model using maximum simulated likelihood. We then extend the technique to a stochastic frontier model with sample selection. In an application that seems superficially obvious, the method is used to revisit the World Health Organization data (WHO in The World Health Report, WHO, Geneva 2000; Tandon et al. in Measuring the overall health system performance for 191 countries, World Health Organization, 2000) where the sample partitioning is based on OECD membership. The original study pooled all 191 countries. The OECD members appear to be discretely different from the rest of the sample. We examine the difference in a sample selection framework.

In 1992, in a dramatic address to the American Association of African Studies, the pioneering figure in the study of African oral traditions, Jan Vansina, identified what he considered to be a major challenge to African history which, if not met, would condemn African history to mediocrity and irrelevance in Africa itself.1 It consisted of two parts: the wholesale transfer of European concepts into Africa around issues of no concern to Africans and a postmodernist attack on historical methodology. Remarking on the ‘irrelevance’ of Luise White’s study of prostitution in Nairobi when Kenyan historians were involved in heated debate about the Mau-Mau, Vansina also criticised White’s interview techniques and what he termed her’ sample selection’, as well as the sampling done by Belinda Bozzoli in her study, Women of Phokeng. Citing instances of ‘hasty’ and’ shoddy’ work he went on at length about the neglect of methodology. Vansina repeatedly addressed the work of David William Cohen, depicting it too as irrelevant, and as a profoundly foreign mode of inquiry2 that both abandons the existing rules for the interpretation of evidence and eschews the possibility of historical truth. In 1995 an amplified version of Vansina’s address appeared as a review article,’ some Perceptions on the Writing of African History, 1948–1992’3 The piece revealed more of the thinking that lay behind the seemingly gratuitous criticisms voiced at the annual meeting three years earlier.

Mortality Selection and Sample Selection: A Comment on Beckett* ANDREW NOYMER University of CaIifornia—BerkeIey Journal of Health and Social Behavior 2001, Vol 42 (Septmber): 326-327 In an interesting article, Megan Beckett (2000) examines the important question of converging health inequalities in later life. Many studies have shown that the differences in health across socioeconomic strata narrow at older ages. Using panel data from the National Health and Nutrition Examination Survey (NHANES), Beckett shows that the converging health inequality cannot be accounted for by mortality selection. The pre- sent comment reconsiders Beckett’s approach to the selection problem, which, while creative, is open to multiple interpretations. Consider the phenomenon that would cause converging health inequalities at later ages to be an “artifact,” as Beckett puts it, of mortality selection. At younger ages, persons with high- er socioeconomic status (SES) have lower lev- els of health problems than those with lower SES. At older ages, the prevalence of health problems in the two groups is closer to parity. If patterns in morbidity are mirrored in mortal- ity, then at older ages a lower SES cohort (higher morbidity and mortality) will be small- er compared to its starting size than a higher SES cohort (lower morbidity and mortality). Since the seminal work of Vaupel, Manton, and Stallard (1979) and Keyfitz and Littman (1979), many demographers have assumed that there are different rates of “frailty” within a population, which determine an individual’s deviation, at any age, from some baseline mor- tality risk. According to the frailty hypothesis, those who die at young ages tend to have high frailty, which skews the distribution of sur- vivors to be more robust. If this condition of nonrandom mortality risks is met, then the aged low SES cohort will be more robust than ‘Address correspondence to: Andrew Noymer, Departments of Sociology and Demography, University of Califomia-—Berkeley, 2232 Pied- mont Avenue, Berkeley, CA 94720; email: andrew@demog.berkeley.edu. when it started out. This reversal of fortune over the life course is called “cohort inversion” (Hobcraft, Menken, and Preston 1982). On the other hand, the low mortality, high SES cohort will have a much less-changed frailty distribu- tion, and will experience less cohort inversion. The greater cohort inversion of the low SES cohort could be enough to overcome the mor- tality disadvantage of being low SES. This problem must be analyzed cautiously, however, as the entire framework for understanding mortality selection effects rests on a counter- factual foundation. If we hold that convergence is a result of mortality selection, we imply that an intrinsic differential persists into older ages and that we would observe it were it not for the selection efiect. On the other hand, if we hold that the convergence is either intrinsic or the result of, for example, access to Medicare (Beckett 2000), we posit that even without selection we would see convergence. In both cases, there is the troubling verb “would.” In reality, we can only observe vital rates that do occur, not those that would occur if some con- dition is met. The general problem of sample selection is encountered frequently in the social sciences (cf. Stolzenberg and Relles 1997; Winship and Mare 1992), and as Beckett (2000) notes, dif- ferential mortality is just a special case of the more general problem. Although we cannot simply “control for” (i.e., condition on) selec- tion bias the same way we would a confound- ing variable, statistical techniques do exist that try to counteract the bias. However, mortality selection is a very special case of sample selec- tion, all the more so if the dependent variable in question is itself health-related. Because of cohort inversion, sample selection due to mor- tality has causal implications beyond nonran- dom missing data in panel followups. This is what makes Beckett’s approach problematic. Consider the statistical technique used by Beckett to set up the hypothetical of no mor- 326

Video annotation is an effective way to facilitate content-based analysis for videos. Automatic machine learning methods are commonly used to accomplish this task. Among these, active learning is one of the most effective methods, especially when the training data cost a great deal to obtain. One of the most challenging problems in active learning is the sample selection. Various sampling strategies can be used, such as uncertainty, density, and diversity, but it is difficult to strike a balance among them. In this paper, we provide a visualization-based batch mode sampling method to handle such a problem. An iso-contour-based scatterplot is used to provide intuitive clues for the representativeness and informativeness of samples and assist users in sample selection. A semisupervised metric learning method is incorporated to help generate an effective scatterplot reflecting the high-level semantic similarity for visual sample selection. Moreover, both quantitative and qualitative evaluations are provided to show that the visualization-based method can effectively enhance sample selection in active learning.

We propose a novel sample selection method for image classification in the presence of noisy labels. Existing methods typically consider small-loss samples as correctly labeled. However, some correctly labeled samples are inherently difficult for the model to learn and can exhibit high loss similar to mislabeled samples in the early stages of training. Consequently, setting a threshold on per-sample loss to select correct labels results in a trade-off between precision and recall in sample selection: a lower threshold may miss many correctly labeled hard-to-learn samples (low recall), while a higher threshold may include many mislabeled samples (low precision). To address this issue, our goal is to accurately distinguish correctly labeled yet hard-to-learn samples from mislabeled ones, thus alleviating the trade-off dilemma. We achieve this by considering the trends in model prediction confidence rather than relying solely on loss values. Empirical observations show that only for correctly labeled samples, the model's prediction confidence for the annotated labels typically increases faster than for any other classes. Based on this insight, we propose tracking the confidence gaps between the annotated labels and other classes during training and evaluating their trends using the Mann-Kendall Test. A sample is considered potentially correctly labeled if all its confidence gaps tend to increase. Our method functions as a plug-and-play component that can be seamlessly integrated into existing sample selection techniques. Experiments on several standard benchmarks and real-world datasets demonstrate that our method enhances the performance of existing methods for learning with noisy labels.

This paper proposes the classification framework based on the Bayesian theory with the single polarization multi-temporal synthetic aperture radar (SAR) and an optical data, and incorporates the proposed training sample selection (SS) methods. Within this framework, the combination with gray level co-occurrence matrix (GLCM)-based mean textural measure is investigated. The two procedures of the classification and proposed SS are united, where SS generates accurate and dispersed training samples. Extracted features from multi-temporal SAR data-namely, the average backscattering coefficient, backscatter temporal variability, and long-term coherence and reflectance values from optical data, are integrated with GLCM mean textural data in the framework. Classification results were generated by taking Osaka City, Japan, as the study area. The selected major classes were built-up areas, fields, woodlands, and water bodies. The most suitable data used for classification is the multi-temporal SAR and an optical data fusion with the mean textural, because of the supplement of different data used and the smoothing effect on the images of the texture. Moreover, the higher quality training samples obtained by using the combined SVM and NN-based SS method for training the Bayesian classifier generated the highest classification accuracies in all of tested cases.

This is the first Quarterly Progress Report for ERDA Contract No. E(49-18)-2475, ''Developing/Modifying Coal Grinding Procedures and Equipment to Produce Predictable Size Distributions During Coal Preparation''. It covers the period, 20 September to 31 December 1976. The project is broken down into 4 Tasks--Study Phase, Preparation Phase, Testing Phase, and Reporting Phase. During this research period, work has been conducted mainly on the Study Phase and the Preparation Phase. Work under the Study Phase has been involved with the subtasks--State of the Art Study, Selection of Samples and Laboratory Analysis. The State of the Art Study has emphasized collection of data from manufacturers of coal size-reduction equipment. The Selection of Samples has selected two coal sources out of the three geographically distinct coals. The Laboratory Analysis has centered around the determination of the primary progeny fragment distribution values for various ranks of coal. Work under the Preparation Phase had been solely aimed at preparing the pilot grinding facility. Work during the next quarter will again be under the Study Phase and Preparation Phase.

As natural language interfaces enable users to express increasingly complex natural language queries, there is a parallel explosion of user review content that can allow users to better find items such as restaurants, books, or movies that match these expressive queries. While Neural Information Retrieval (IR) methods have provided state-of-the-art results for matching queries to documents, they have not been extended to the task of Reviewed-Item Retrieval (RIR), where query-review scores must be aggregated (or fused) into item-level scores for ranking. In the absence of labeled RIR datasets, we extend Neural IR methodology to RIR by leveraging self-supervised methods for contrastive learning of BERT embeddings for both queries and reviews. Specifically, contrastive learning requires a choice of positive and negative samples, where the unique two-level structure of our item-review data combined with meta-data affords us a rich structure for the selection of these samples. For contrastive learning in a Late Fusion scenario (where we aggregate query-review scores into item-level scores), we investigate the use of positive review samples from the same item and/or with the same rating, selection of hard positive samples by choosing the least similar reviews from the same anchor item, and selection of hard negative samples by choosing the most similar reviews from different items. We also explore anchor sub-sampling and augmenting with meta-data. For a more end-to-end Early Fusion approach, we introduce contrastive item embedding learning to fuse reviews into single item embeddings. Experimental results show that Late Fusion contrastive learning for Neural RIR outperforms all other contrastive IR configurations, Neural IR, and sparse retrieval baselines, thus demonstrating the power of exploiting the two-level structure in Neural RIR approaches as well as the importance of preserving the nuance of individual review content via Late Fusion methods.

Up-to-date and reliable information on land cover and land use status is important in many aspects of human activities. Knowledge about the reference dataset, its coverage, nomenclature, thematic and geometric accuracy, spatial resolution is crucial for appropriate selection of reference samples used in the classification process. In this study, we examined the impact of the selection and pre-processing of reference samples for the classification accuracy. The classification based on Random Forest algorithm was performed using firstly the automatically selected reference samples derived directly from the national databases, and secondly using the pre-processed and verified reference samples. The verification procedures involved the iterative analysis of histogram of spectral features derived from the Sentinel-2 data for individual land cover classes. The verification of the reference samples improved the accuracy of delineation of all land cover classes. The highest improvement was achieved for the woodland broadleaved and non- and sparce vegetation classes, with the overall accuracy increasing from 51% to 73%, and from 33% to 74%, respectively. The second objective of this study was to derive the best possible land cover classification over the mountain area in Norway, therefore we examined whether the use of the Digital Elevation Model (DEM) can improve the classification results. Classifications were carried out based on Sentinel-2 data and a combination of Sentinel-2 and DEM. Using the DEM the accuracy for nine out of ten land cover classes was improved. The highest improvement was achieved for classes located at higher altitudes: low vegetation and non- and sparse vegetation.

Supervised classification is the commonly used method for extracting ground information from images. However, for supervised classification, the selection and labelling of training samples is an expensive and time-consuming task. Recently, automatic information indexes have achieved satisfactory results for indicating different land-cover classes, which makes it possible to develop an automatic method for labelling the training samples instead of manual interpretation. In this paper, we propose a method for the automatic selection and labelling of training samples for high-resolution image classification. In this way, the initial candidate training samples can be provided by the information indexes and open-source geographical information system (GIS) data, referring to the representative land-cover classes: buildings, roads, soil, water, shadow, and vegetation. Several operations are then applied to refine the initial samples, including removing overlaps, removing borders, and semantic constraints. The proposed sampling method is evaluated on a series of high-resolution remote sensing images over urban areas, and is compared to classification with manually labeled training samples. It is found that the proposed method is able to provide and label a large number of reliable samples, and can achieve satisfactory results for different classifiers. In addition, our experiments show that active learning can further enhance the classification performance, as active learning is used to choose the most informative samples from the automatically labeled samples.

The purpose of this study is to investigate the influence of financial leverage, liquidity, asset growth, size on the beta stock in manufacturing companies in Indonesia Stock Exchange. The sample in this study is manufacturing companies listed in Indonesia Stock Exchange during the 5 periods (2008-2012). Total samples are 58 companies. The sample selection is done by using the method of stratified random sampling. The multiple linear regression analysis was used to test the hypothesis.This research found that asset growth has significant effect on systematic risk, while variable financial leverage, liquidity, sizedid not have significant effect on systematic risk. The purpose of this study is to investigate the influence of financial leverage, liquidity, asset growth, sizeon the beta stock in manufacturing companies in Indonesia Stock Exchange. The sample in this study is manufacturing companies listed in Indonesia Stock Exchange during the 5 periods (2008-2012). Total samples are 58 companies. The sample selection is done by using the method of stratified random sampling. The multiple linear regression analysis was used to test the hypothesis.This research found that asset growth has significant effect on systematic risk, while variable financial leverage, liquidity, sizedid not have significant effect on systematic risk.

This paper explores the possibility of constructing RBF classifiers which, somewhat like support vector machines, use a reduced number of samples as centroids, by means of selecting samples in a direct way. Because sample selection is viewed as a hard computational problem, this selection is done after a previous vector quantization: this way obtaining also other similar machines using centroids selected from those that are learned in a supervised manner. Several forms of designing these machines are considered, in particular with respect to sample selection; as well as some different criteria to train them. Simulation results for well-known classification problems show very good performance of the corresponding designs, improving that of support vector machines and reducing substantially their number of units. This shows that our interest in selecting samples (or centroids) in an efficient manner is justified. Many new research avenues appear from these experiments and discussions, as suggested in our conclusions.

SummaryWe consider estimating binary response models on an unbalanced panel, where the outcome of the dependent variable may be missing due to nonrandom selection, or there is self‐selection into a treatment. In the present paper, we first consider estimation of sample selection models and treatment effects using a fully parametric approach, where the error distribution is assumed to be normal in both primary and selection equations. Arbitrary time dependence in errors is permitted. Estimation of both coefficients and partial effects, as well as tests for selection bias, are discussed. Furthermore, we consider a semiparametric estimator of binary response panel data models with sample selection that is robust to a variety of error distributions. The estimator employs a control function approach to account for endogenous selection and permits consistent estimation of scaled coefficients and relative effects.

PurposeThe rich financial knowledge of small and micro business owners helps to enhance the formal credit demand of small and micro enterprises and change the credit channel preference of small and micro enterprises. The purpose of this paper is to explore the relationship between financial knowledge and the credit practices of 290 small and micro enterprises in China’s Jiangsu and Shandong provinces based on their formal credit needs and preferred channels of credit.Design/methodology/approachTo measure the degree of the credit constraints of small and micro enterprises, this study applied questionnaire surveys to obtain information on the credit demand and supply of 363 small and micro enterprises in the Jiangsu and Shandong provinces. Firstly, a probit model is used to study the influence of financial knowledge on the formal credit demand and credit acquisition possibility of small and micro enterprises, and tool variables and a biprobit model are used to deal with the possible errors of endogenesis and sample selection. Secondly, a tobit model is used to study the influence of financial knowledge on the credit access of small and micro enterprises in different channels, and tool variables and a Heckman two-stage model are used to deal with endogenesis and possible errors in sample selection. Finally, this study carried out a series of robustness tests to make the conclusions more reliable.FindingsThis study is based on the perspective of the knowledge-based view to explore the impact of financial knowledge on the credit behaviour of small and micro enterprises. This study found that financial knowledge can increase a small and micro enterprise’s formal credit needs and drive the small and micro enterprise to actively apply for loans. Furthermore, financial knowledge has a significant and positive influence on the acquisition of formal credit and approved lines of formal credit and a significant and negative influence on the acquisition of informal credit and approved lines of informal credit.Research limitations/implicationsThe results indicated that increased financial knowledge can increase the likelihood of a small and micro enterprise to prefer formal credit and reduce the likelihood of it to prefer informal credit channels.Originality/valueFinancial knowledge is the ability to master basic economic knowledge and financial concepts as well as the ability to use knowledge to manage and allocate financial resources. The rich financial knowledge of small and micro business owners helps to enhance the formal credit demand of small and micro enterprises and change their credit channel preference. This paper offers a new perspective on the problems of credit constraint, low participation in formal credit markets and high participation in private credit markets among China’s small and micro enterprises and valuably supplements the research literature.