Main Dataset Research Articles

Agricultural loan delinquency prediction using machine learning methods

The recent economic downturn in the agricultural sector that started in 2013 has caused some concerns for farmers’ repayment capacity, which raises the need for precise prediction of financial stress in the agricultural sector. Machine learning has been shown to improve predictions with large financial data, however, its application remains limited in the agricultural sector. In this study, we approximate financial stress by agricultural loan delinquency, and predict it by employing a logistic regression and several machine learning methods. The main datasets include the Call Reports and Summary of Deposits from the Federal Deposit Insurance Corporation (FDIC). Our results show that ensemble learning methods have the best performance in prediction accuracy, with improvement of 26 percentage points at most and that the Naïve Bayes classifier is the best method to maintain the lowest cost from false predictions when the failure of identifying potentially high-risk loans is very costly. From the perspective of banks, while there are important benefits to using machine learning, the bank-level costs are also important considerations that may lead to different choices of machine learning methods.

Integrated water vapour observations in the Caribbean arc from a network of ground-based GNSS receivers during EUREC<sup>4</sup>A

Abstract. Ground-based Global Navigation Satellite System (GNSS) measurements from nearly 50 stations distributed over the Caribbean arc have been analysed for the period 1 January–29 February 2020 in the framework of the EUREC4A (Elucidate the Couplings Between Clouds, Convection and Circulation) field campaign. The aim of this effort is to deliver high-quality integrated water vapour (IWV) estimates to investigate the moisture environment of mesoscale cloud patterns in the trade winds and their feedback on the large-scale circulation and energy budget. This paper describes the GNSS data processing procedures and assesses the quality of the GNSS IWV retrievals from four operational streams and one reprocessed research stream which is the main data set used for offline scientific applications. The uncertainties associated with each of the data sets, including the zenith tropospheric delay (ZTD)-to-IWV conversion methods and auxiliary data, are quantified and discussed. The IWV estimates from the reprocessed data set are compared to the Vaisala RS41 radiosonde measurements operated from the Barbados Cloud Observatory (BCO) and to the measurements from the operational radiosonde station at Grantley Adams International Airport (GAIA), Bridgetown, Barbados. A significant dry bias is found in the GAIA humidity observations with respect to the BCO sondes (−2.9 kg m−2) and the GNSS results (−1.2 kg m−2). A systematic bias between the BCO sondes and GNSS is also observed (1.7 kg m−2), where the Vaisala RS41 measurements are moister than the GNSS retrievals. The IWV estimates from a collocated microwave radiometer agree with the BCO soundings after an instrumental update on 27 January, while they exhibit a dry bias compared to the soundings and to GNSS before that date. IWV estimates from the ECMWF fifth-generation reanalysis (ERA5) are overall close to the GAIA observations, probably due to the assimilation of these observations in the reanalysis. However, during several events where strong peaks in IWV occurred, ERA5 is shown to significantly underestimate the GNSS-derived IWV peaks. Two successive peaks are observed on 22 January and 23–24 January which were associated with heavy rain and deep moist layers extending from the surface up to altitudes of 3.5 and 5 km, respectively. ERA5 significantly underestimates the moisture content in the upper part of these layers. The origins of the various moisture biases are currently being investigated. We classified the cloud organization for five representative GNSS stations across the Caribbean arc using visible satellite images. A statistically significant link was found between the cloud patterns and the local IWV observations from the GNSS sites as well as the larger-scale IWV patterns from the ECMWF ERA5 reanalysis. The reprocessed ZTD and IWV data sets from 49 GNSS stations used in this study are available from the French data and service centre for atmosphere (AERIS) (https://doi.org/10.25326/79; Bock, 2020b).

Monocular multi-person pose estimation: A survey

Multi-person pose estimation in unconstrained scenarios, with an unknown number of individuals, is a main step towards scene understanding and action recognition. Due to the recent advancements on the architecture of convolutional networks, body part detectors are now accurate and estimate poses in real-time for both single- and multi-person scenes. In contrast, assigning detected body parts to coherent human poses when there are multiple persons interacting is an arduous task. To name a few of the challenges that arise in such scenes: person-to-person occlusion, truncated body parts, and more sources for double counting. Recently, the community contributed towards solving most of them. Hence, it would be interesting to analyze and compile successful approaches from current literature into research trends, and identify possible gaps for future works. To the best of our knowledge, there is no up-to-date review on the main advancements in the field that target this particular set of challenges. This survey fills this gap by reviewing the main breakthroughs on multi-person pose estimation over the last decade and summarizing their impact on the state-of-the-art. Regarding scientific contributions, we propose a novel taxonomy that categorizes the reviewed methods according to their main contributions to the pose estimation pipeline, lists the main datasets and evaluation metrics to train new models, and provides insights on the best entries of publicly available benchmarks.

Are We Hungry for 3D LiDAR Data for Semantic Segmentation? A Survey of Datasets and Methods

3D semantic segmentation is a fundamental task for robotic and autonomous driving applications. Recent works have been focused on using deep learning techniques, whereas developing fine-annotated 3D LiDAR datasets is extremely labor intensive and requires professional skills. The performance limitation caused by insufficient datasets is called data hunger problem. This research provides a comprehensive survey on the question: are we hungry for 3D LiDAR data for semantic segmentation? The studies are conducted at three levels. First, a broad review to the main 3D LiDAR datasets is conducted, followed by a statistical analysis on three representative datasets to gain an in-depth view on the datasets’ size, diversity and quality, which are the critical factors in learning deep models. Second, an organized survey of 3D semantic segmentation methods is given with a focus on the mainstream of the latest research trend using deep learning techniques, followed by a systematic survey to the existing efforts to solve the data hunger problem. Finally, an insightful discussion of the remaining problems on both methodological and datasets’ viewpoints, and the open questions on dataset bias, domain and semantic gap are given, leading to potential topics in future works. To the best of our knowledge, this is the first work to study the data hunger problem for 3D semantic segmentation using deep learning techniques, which are addressed in both methodological and dataset review, and we share findings and discussions through a comprehensive dataset analysis.

Blood lead level monitoring related to environmental exposure in the general Iranian population: a systematic review and meta-analysis.

Exposure to lead can cause adverse health problems incorporating hypophosphatemia, heart and liver disease, cancers, neurological and cardiovascular diseases, central nervous disorders, and sensory disorders. This study investigated the blood lead level in the general Iranian population with environmental exposure to lead. In the presented systematic review and meta-analysis, the authors searched Iranian dataset, including Magiran, SID, Iranmedex, and Nopa, and the main dataset, comprising PubMed, Scopus, Embase, and Web of Science, all available articles until 12 January 2019, and extracting 55 studies (with 63 data for analysis) to a meta-analysis. A comprehensive meta-analysis software, pooled standard deviation, mean, sample size, and the utterly random effects model was analyzed in this study. The results showed that the overall mean BLL (95% CI) in total inquiries was 6.41 μg/dL (5.96 to 6.87). Besides, the results for gender and age subgroups were as follows: mean BLL, 6.47 μg/dL, 95% CI, 5.79, to 7.15, mean BLL, 6.44 μg/dL, 95% CI, 5.96, to 6.91, respectively. Conclusively the mean BLL in the Iranian population was higher than the recommended level by the US Centers for Disease Control and Prevention (CDC). Results indicated that the mean BLL in men and adults was more elevated than in women and children, respectively. Therefore, BLL monitoring and screening of the general Iranian population are is necessary to determine a reference value.

110 Physiological correlates of the Epworth Sleepiness Scale reveal different dimensions of daytime sleepiness

Abstract Introduction The Epworth Sleepiness Scale (ESS) is used as a clinical tool for determining excessive daytime sleepiness. However, the behavior and biology that underlie ESS scores remain to be elucidated. The main objective of this analysis is to determine objective behavioral and physiologic correlates of the ESS. Secondarily, we examine the relationship of the ESS to parallel subjective and objective endpoints that could represent measures of daytime sleepiness. Methods Using two separate machine learning algorithms, Random Forest and Lasso, we determined the association between ESS scores and 55 sleep and medical variables in individuals who participated in the Sleep Heart Health Study (N=2105). These variables include self-reported sleep characteristics (e.g., habitual sleep length and latency, frequency of not getting enough sleep), polysomnographic sleep measures from a single night, medication use, and mental and physical health status. Additional analyses were conducted on data stratified by age and gender. To investigate the relationship between ESS and other measures of daytime sleepiness, cross-correlation analysis was conducted on the ESS and five variables that could analog daytime sleepiness (feeling unrested, nap duration and frequency, sleep latency, frequency of not getting enough sleep). Results Analysis of the main dataset resulted in low explained variance (7.15 - 10.0%), with self-reported frequency of not getting enough sleep as most important predictor (10.3–13.9% of the model variance). Stratification by neither age nor gender significantly improved explained variance. Habitual sleep length was not an important predictor in any model. Cross-correlational analysis revealed low correlation of other daytime sleepiness measures to ESS score. Conclusion Data analyses indicate that ESS scores are not well explained by habitual or polysomnographic sleep values, or a variety of other biomedical characteristics. This suggests that there are different, potentially orthogonal dimensions of the concept of “daytime sleepiness” that may be driven by different aspects of sleep physiology. Caution should be used when considering the ESS as a clinical measure given that the physiologic correlates still remain to be elucidated. Support (if any):

The role of electricity mix and production efficiency improvements on greenhouse gas (GHG) emissions of building components and future refurbishment measures

PurposeAn estimation of the environmental impact of buildings by means of a life cycle assessment (LCA) raises uncertainty related to the parameters that are subject to major changes over longer time spans. The main aim of the present study is to evaluate the influence of modifications in the electricity mix and the production efficiency in the chosen reference year on the embodied impacts (i.e., greenhouse gas (GHG) emissions) of building materials and components and the possible impact of this on future refurbishment measures.MethodsA new LCA methodological approach was developed and implemented that can have a significant impact on the way in which existing buildings are assessed at the end of their service lives. The electricity mixes of different reference years were collected and assessed, and the main datasets and sub-datasets were modified according to the predefined substitution criteria. The influence of the electricity-mix modification and production efficiency were illustrated on a selected existing reference building, built in 1970. The relative contribution of the electricity mix to the embodied impact of the production phase was calculated for four different electricity mixes, with this comprising the electricity mix from 1970, the current electricity mix and two possible future electricity-mix scenarios for 2050. The residual value of the building was also estimated.Results and discussionIn the case presented, the relative share of the electricity mix GHG emission towards the total value was as high as 20% for separate building components. If this electricity mix is replaced with an electricity mix having greater environmental emissions, the relative contribution of the electricity mix to the total emissions can be even higher. When, by contrast, the modified electricity mix is almost decarbonized, the relative contribution to the total emissions may well be reduced to a point where it becomes negligible. The modification of the electricity mix can also influence the residual value of a building. In the observed case, the differences due to different electricity mixes were in the range of 10%.ConclusionsIt was found that those parameters that are subject to a major change during the reference service period of the building should be treated dynamically in order to obtain reliable results. Future research is foreseen to provide additional knowledge concerning the influence of dynamic parameters on both the use phase and the end-of-life phase of buildings, and these findings will also be important when planning future refurbishment measures.

Countering Spoof: Towards Detecting Deepfake with Multidimensional Biological Signals

The deepfake technology is conveniently abused with the low technology threshold, which may bring the huge social security risks. As GAN-based synthesis technology is becoming stronger, various methods are difficult to classify the fake content effectively. However, although the fake content generated by GANs can deceive the human eyes, it ignores the biological signals hidden in the face video. In this paper, we proposed a novel video forensics method with multidimensional biological signals, which extracting the difference of the biological signal between real and fake videos from three dimensions. The experimental results show that our method achieves 98% accuracy on the main public dataset. Compared with other technologies, the proposed method only extracts fake video information and is not limited to a specific generation method, so it is not affected by synthetic methods and has good adaptability.

Eye-color and Type-2 diabetes phenotype prediction from genotype data using deep learning methods

BackgroundGenotype–phenotype predictions are of great importance in genetics. These predictions can help to find genetic mutations causing variations in human beings. There are many approaches for finding the association which can be broadly categorized into two classes, statistical techniques, and machine learning. Statistical techniques are good for finding the actual SNPs causing variation where Machine Learning techniques are good where we just want to classify the people into different categories. In this article, we examined the Eye-color and Type-2 diabetes phenotype. The proposed technique is a hybrid approach consisting of some parts from statistical techniques and remaining from Machine learning.ResultsThe main dataset for Eye-color phenotype consists of 806 people. 404 people have Blue-Green eyes where 402 people have Brown eyes. After preprocessing we generated 8 different datasets, containing different numbers of SNPs, using the mutation difference and thresholding at individual SNP. We calculated three types of mutation at each SNP no mutation, partial mutation, and full mutation. After that data is transformed for machine learning algorithms. We used about 9 classifiers, RandomForest, Extreme Gradient boosting, ANN, LSTM, GRU, BILSTM, 1DCNN, ensembles of ANN, and ensembles of LSTM which gave the best accuracy of 0.91, 0.9286, 0.945, 0.94, 0.94, 0.92, 0.95, and 0.96% respectively. Stacked ensembles of LSTM outperformed other algorithms for 1560 SNPs with an overall accuracy of 0.96, AUC = 0.98 for brown eyes, and AUC = 0.97 for Blue-Green eyes. The main dataset for Type-2 diabetes consists of 107 people where 30 people are classified as cases and 74 people as controls. We used different linear threshold to find the optimal number of SNPs for classification. The final model gave an accuracy of 0.97%.ConclusionGenotype–phenotype predictions are very useful especially in forensic. These predictions can help to identify SNP variant association with traits and diseases. Given more datasets, machine learning model predictions can be increased. Moreover, the non-linearity in the Machine learning model and the combination of SNPs Mutations while training the model increases the prediction. We considered binary classification problems but the proposed approach can be extended to multi-class classification.

Double Exponential Smoothing Berimputasi LOCF Dan Linear Interpolation Dalam Akurasi Peramalan Harga Harian Emas

Gold is another kind of investment that often experiences price change, mostly every day. Because of its price fluctuation, forecasting is needed to help the investor in investment decision making. But during this Coronavirus Disease 2019 (Covid-19), the gold price is fluctuating extremely than the past 4 years so a better forecasting method approached and analysis technique is needed due to this case. Double Exponential Smoothing method is chosen to forecast this daily gold price. On the other hand, there are so many missing values spreading around the main dataset so the imputation method is needed too, Last Observation Carried Forward (LOCF) and linear interpolation are chosen for imputing the missing values. In this research, the main dataset was split into 3 (three) datasets, which are Precovid-19 (before Covid-19, used only for visualizing the actual fluctuation condition during this pandemic), Incovid-19 (during Covid-19 based on the date where the first Covid-19 case occurred in Indonesia), and Combination (a binding dataset of Pracovid-19 and Incovid-19). Although Incovid-19’s MAPE value is higher than Pracovid-19 and Combination’s MAPE values, in evaluation session showed that Incovid-19’s MAPE of forecast results has the lowest value rather than Combination’s MAPE of forecast results, so the conclusion of this research is Incovid-19 dataset with LOCF imputation is the most adaptive with the actual condition and it is used to forecast the daily gold price until the last period of the main dataset then.

The Impact of COVID-19 on Immigration Detention

COVID-19 has spread quickly through immigration detention facilities in the United States. As of December 2, 2020, there have been over 7,500 confirmed COVID-19 cases among detained noncitizens. This Article examines why COVID-19 spread rapidly in immigration detention facilities, how it has transformed detention and deportation proceedings, and what can be done to improve the situation for detained noncitizens. Part I identifies key factors that contributed to the rapid spread of COVID-19 in immigration detention. While these factors are not an exhaustive list, they highlight important weaknesses in the immigration detention system. Part II then examines how the pandemic changed the size of the population in detention, the length of detention, and the nature of removal proceedings. In Part III, the Article offers recommendations for mitigating the impact of COVID-19 on detained noncitizens. These recommendations include using more alternatives to detention, curtailing transfers between detention facilities, establishing a better tracking system for medically vulnerable detainees, prioritizing bond hearings and habeas petitions, and including immigration detainees among the groups to be offered COVID-19 vaccine in the initial phase of the vaccination program. The lessons learned from the spread of COVID-19 in immigration detention will hopefully lead to a better response to any future pandemics. In discussing these issues, the Article draws on national data from January 2019 through November 2020 published by Immigration and Customs Enforcement (ICE) and Customs and Border Protection (CBP), two agencies within DHS. The main datasets used are detention statistics published by ICE for FY 2019 (Oct. 2018-Sep. 2019), FY 2020 (Oct. 2019-Sep. 2020), and the first two months of FY 2021 (Oct. 2020-Nov. 2020). These datasets include detention statistics about individuals arrested by ICE in the interior of the country, as well as by CBP at or near the border. Additionally, the Article draws on separate data published by CBP regarding the total number of apprehensions at the border based on its immigration authority under Title 8 of the United States Code, as well as the number of expulsions at the border based on its public health authority under Title 42 of the United States Code.

Deep-learning approach in the study of skin lesions.

Medical technology is far from reaching its full potential. An area that is currently expanding is that of precision medicine. The aim of this article is to present an application of precision medicine-a deep-learning approach to computer-aided diagnosis in the field of dermatology. The main dataset was proposed in the edition of the ISIC Challenge that took place in 2019 and included 25331 dermoscopic images from eight different categories of lesions-three of them were malignant and five benign. The behavior of the model was also tested on a dataset collected from the second Department of Dermatology, of the Colentina Clinical Hospital. The overall accuracy of the model was 78.11%. Of the total 5031 samples included in the test subset, 3958 were correctly classified. The accuracy of the model on the clinical dataset is lower than that obtained in the first instance. The architecture of the model can be considered of general use, being able to be adapted in an optimal way for a wide range of classifications. The model has achieved performance within the expected limits but can be further improved by new methods.

Velar lateral allophony in Mee (Ekari)

Velar lateral sounds are rare and their acoustics and contextual effects are understudied. Positional variants of velar laterals are also rarely reported. This paper documents a previously unknown allophony pattern of the velar lateral in Mee (Trans New Guinea; Indonesia), based on an elicitation study with two speakers and a controlled set of recordings from one of the speakers. Our main dataset included carrier phrase recordings of Mee words with the velar lateral, representing the diverse set of contexts where the velar lateral occurs. Our acoustic findings suggest that the Mee velar lateral is realized as a laterally released velar stop [ɡᶫ] before front vowels, and with uvular closure followed by a fricative release [ɢʁ] before back vowels. In line with this description, we found differences in the second formant of the preceding vowel and the periodicity of the release for the two allophones. We explore the implications of our findings for the typology of velar laterals.

Empirical Analyses of Income: Finland (2009) and Australia (1967-1968)

Analyses of income data are often based on assumptions concerning theoretical distributions. In this study, we apply statistical analyses, but ignore specific distribution models. The main income data sets considered in this study are taxable income in Finland (2009) and household income in Australia (1967-1968). Our intention is to compare statistical analyses performed without assumptions of the theoretical models with earlier results based on specific models. We have presented the central objects, probability density function, cumulative distribution function, the Lorenz curve, the derivative of the Lorenz curve, the Gini index and the Pietra index. The trapezium rule, Simpson´s rule, the regression model and the difference quotients yield comparable results for the Finnish data, but for the Australian data the differences are marked. For the Australian data, the discrepancies are caused by limited data. JEL classification numbers: D31, D63, E64. Keywords: Cumulative distribution function, Probability density function, Mean, quantiles, Lorenz curve, Gini coefficient, Pietra index, Robin Hood index, Trapezium rule, Simpson´s rule, Regression models, Difference quotients, Derivative of Lorenz curve

Physiological correlates of the Epworth Sleepiness Scale reveal different dimensions of daytime sleepiness.

The Epworth Sleepiness Scale is commonly used to examine self-reported daytime sleepiness in clinical populations; the physiologic correlates of this scale, however, are not well understood. Furthermore, how well this scale correlates with parallel objective and self-reported concepts of daytime sleepiness is not well described. As such, we used machine learning algorithms to examine the association between Epworth Sleepiness Scale scores and 55 sleep and medical variables in the Sleep Heart Health Study (N = 2105). Secondary analyses examined data stratified by age and gender and the relationship between the Epworth and other measures of daytime sleepiness. Analyses of the main data set resulted in low explained variance (7.15%-10.0%), with self-reported frequency of not getting enough sleep as most important predictor (10.3%-13.9% of the model variance). Stratification by neither age nor gender significantly improved explained variance. Cross-correlational analysis revealed low correlation of other daytime sleepiness measures to Epworth scores. We find that Epworth scores are not well explained by habitual or polysomnographic sleep values, or other biomedical characteristics. These analyses indicate that there are different, potentially orthogonal dimensions of the concept of "daytime sleepiness" that may be driven by different aspects of sleep physiology. As the physiologic correlates of the Epworth Sleepiness Scale remain to be elucidated, interpretation of the clinical meaning of these scores should be done with caution.

Assessing current and future spatiotemporal precipitation variability and trends over Uganda, East Africa, based on CHIRPS and regional climate model datasets

The lack of reliable rainfall projection records remains a major challenge to Uganda. In the advent of extreme wetness or drought events, reliable rainfall estimates for local planning and adaptation are essential. The present study used two main datasets to conduct a historical analysis from 1981 to 2019, coupled with future projections under representative concentration pathway (RCP 8.5) for the period 2020–2050. Historical analysis revealed bimodal annual rainfall patterns for March–May (MAM) and September–November (SON) gradients representing heavier to lighter rainfall events, respectively, over the study area. Investigation of recent trends in rainfall patterns revealed an upward trend from 2010 onwards in annual and seasonal rainfall. Moreover, results for future projections show wet conditions are projected to occur over the study area between the months of April/May and October. Contrarily, March is likely to experience a reduction in wet conditions. Mann–Kendall test employed to make future projections of rainfall depicted decreasing patterns during MAM season whilst increasing tendencies with strong shift was highlighted for SON season over the study region. Meanwhile, annual projections indicate huge variations with linear trends showing a marginal increase as compared to historical trends. Findings would serve as baseline print to propel further studies that could delve into impact analysis of drought extreme events which pose significant threats to the agricultural sector which is heavily reliant on rainfall.

Improved prediction and characterization of anticancer activities of peptides using a novel flexible scoring card method

As anticancer peptides (ACPs) have attracted great interest for cancer treatment, several approaches based on machine learning have been proposed for ACP identification. Although existing methods have afforded high prediction accuracies, however such models are using a large number of descriptors together with complex ensemble approaches that consequently leads to low interpretability and thus poses a challenge for biologists and biochemists. Therefore, it is desirable to develop a simple, interpretable and efficient predictor for accurate ACP identification as well as providing the means for the rational design of new anticancer peptides with promising potential for clinical application. Herein, we propose a novel flexible scoring card method (FSCM) making use of propensity scores of local and global sequential information for the development of a sequence-based ACP predictor (named iACP-FSCM) for improving the prediction accuracy and model interpretability. To the best of our knowledge, iACP-FSCM represents the first sequence-based ACP predictor for rationalizing an in-depth understanding into the molecular basis for the enhancement of anticancer activities of peptides via the use of FSCM-derived propensity scores. The independent testing results showed that the iACP-FSCM provided accuracies of 0.825 and 0.910 as evaluated on the main and alternative datasets, respectively. Results from comparative benchmarking demonstrated that iACP-FSCM could outperform seven other existing ACP predictors with marked improvements of 7% and 17% for accuracy and MCC, respectively, on the main dataset. Furthermore, the iACP-FSCM (0.910) achieved very comparable results to that of the state-of-the-art ensemble model AntiCP2.0 (0.920) as evaluated on the alternative dataset. Comparative results demonstrated that iACP-FSCM was the most suitable choice for ACP identification and characterization considering its simplicity, interpretability and generalizability. It is highly anticipated that the iACP-FSCM may be a robust tool for the rapid screening and identification of promising ACPs for clinical use.

Evaluating equality in prescribing Novel Oral Anticoagulants (NOACs) in England: The protocol of a Bayesian small area analysis

Atrial fibrillation (AF) is the most common cardiac arrhythmia, affecting about 1.6% of the population in England. Novel oral anticoagulants (NOACs) are approved AF treatments that reduce stroke risk. In this study, we estimate the equality in individual NOAC prescriptions with high spatial resolution in Clinical Commissioning Groups (CCGs) across England from 2014 to 2019. A Bayesian spatio-temporal model will be used to estimate and predict the individual NOAC prescription trend on 'prescription data' as an indicator of health services utilisation, using a small area analysis methodology. The main dataset in this study is the "Practice Level Prescribing in England," which contains four individual NOACs prescribed by all registered GP practices in England. We will use the defined daily dose (DDD) equivalent methodology, as recommended by the World Health Organization (WHO), to compare across space and time. Four licensed NOACs datasets will be summed per 1,000 patients at the CCG-level over time. We will also adjust for CCG-level covariates, such as demographic data, Multiple Deprivation Index, and rural-urban classification. We aim to employ the extended BYM2 model (space-time model) using the RStan package. This study suggests a new statistical modelling approach to link prescription and socioeconomic data to model pharmacoepidemiologic data. Quantifying space and time differences will allow for the evaluation of inequalities in the prescription of NOACs. The methodology will help develop geographically targeted public health interventions, campaigns, audits, or guidelines to improve areas of low prescription. This approach can be used for other medications, especially those used for chronic diseases that must be monitored over time.

An enhanced siamese angular softmax network with dual joint-attention for person re-identification

For person re-identification (re-ID), a core problem is how to learn discriminative feature representations of pedestrians. In this paper, we propose a novel enhanced siamese angular softmax network (ES-ASnet) to integrate identification, verification and metric learning into a unified network. First, a dual joint-attention (DJA) based identification model is proposed that can focus on both key local information and global contextual dependencies in spatial and channel domains simultaneously. Then, we adopt angular softmax (A-Softmax) loss in the training phase, which directly integrates metric learning into classification to enhance the discriminative capability of features in the angular space. Furthermore, the alignment module in the unified network can reduce the impact of misalignment between image pairs, which can further learn robust discriminative feature representations effectively. Experiments on three main person re-ID datasets, including Market1501, DukeMTMC-reID and CUHK03-NP, demonstrate that the proposed network has achieved competitive performance compared with several state-of-the-art methods for person re-ID.

The Construct of Sex Work in France: A Qualitative Analysis of the Development of a Sex Work Policy

This paper aims to shed light on the particular development of repressive policies on sex work in France. The analysis draws on two main datasets: ongoing research on humanitarian interventions and interviews with insti-tutional, mainstream feminists and other neo-abolitionist groups in Paris and Marseille. The analysis is supported by the ethnographic data emerging from interviews with street- and internet-based migrant workers and key informants, such as community health outreach workers, social workers, and immigration lawyers, in Paris. This analysis argues that the construction of prostitution as a social problem ought to be seen in light of broader political anxieties over sexist behaviour in poor neighbourhoods and immigration control.