Explore the vast range of titles available.

The increasingly common use of neural network classifiers in industrial and social applications of image analysis has allowed impressive progress these last years. Such methods are, however, sensitive to algorithmic bias, i.e., to an under- or an over-representation of positive predictions or to higher prediction errors in specific subgroups of images. We then introduce in this paper a new method to temper the algorithmic bias in Neural-Network-based classifiers. Our method is Neural-Network architecture agnostic and scales well to massive training sets of images. It indeed only overloads the loss function with a Wasserstein-2-based regularization term for which we back-propagate the impact of specific output predictions using a new model, based on the Gâteaux derivatives of the predictions distribution. This model is algorithmically reasonable and makes it possible to use our regularized loss with standard stochastic gradient-descent strategies. Its good behavior is assessed on the reference Adult census , MNIST , CelebA datasets.

Linear programs with quadratic (“ridge”) regularization are of recent interest in optimal transport: unlike entropic regularization, the squared-norm penalty gives rise to sparse approximations of optimal transport couplings. More broadly, quadratic regularization is used in overparametrized learning problems to single out a particular solution. It is well known that the solution of a quadratically regularized linear program over any polytope converges stationarily to the minimal-norm solution of the linear program when the regularization parameter tends to zero. However, that result is merely qualitative. Our main result quantifies the convergence by specifying the exact threshold for the regularization parameter, after which the regularized solution also solves the linear program. Moreover, we bound the suboptimality of the regularized solution before the threshold. These results are complemented by a convergence rate for the regime of large regularization. We apply our general results to the setting of optimal transport, where we shed light on how the threshold and suboptimality depend on the number of data points.

There are many cropping systems on floodplain soils, but greenhouse gas (GHG) emission balances of these agricultural systems are rarely reported. Carbon (C) footprints of agricultural products were assessed using a co-designed life cycle assessment tool in major cropping systems in Bangladesh: rice-rice-rice (R-R-R/boro-aus-aman), rice-fallow-rice (R-F-R/boro-fallow-aman), maize-fallow-rice (M-F-R), wheat-mungbean-rice (W-Mu-R), and potato-rice-fallow (P-R-F) along with the field measurement of some of the systems. The rice system with dryland crops had higher nitrous oxide (N 2 O) emissions (3.8 in maize, 4.5 in potato and 0.92 kg N 2 O–N ha −1 in mungbean) than sole rice (0.73 in boro, 0.57 in aus and 1.94 kg N 2 O–N ha −1 in aman) systems but methane (CH 4 ) emissions exhibited the opposite. Methane dominated, accounting for 50–80% of total emissions in rice systems. The boro rice-based systems (R-R-R and R-F-R) had the highest C footprint (ca. 25.8 and 19.2 Mg CO 2 e ha −1 ) while the P-F-R (12.3 Mg CO 2 e ha −1 ) and M-F-R (12.6 Mg CO 2 e ha −1 ) had the lowest C footprint. Boro and aus were more suitable to reduce C footprint. Measured CH 4 and N 2 O data agreed well with the IPCC Tier 1 estimates but further study on GHG measurements in other agroecosystems and cropping systems are required to validate the estimation for adopting suitable GHG mitigation strategies.

Ensuring food production with low nitrogen (N) environmental emissions requires good quantitative knowledge of N flows in agricultural systems to monitor emissions and N use efficiency (NUE, the ratio of N outputs to inputs). Our study quantifies the main N agricultural flows at subnational and national scales in Denmark and Sweden from 2011 to 2020, calculating the NUE for crop and livestock production and associated nitrous oxide (N2O) and ammonia (NH3) emissions. In Denmark, our results indicate a similar use of organic (manure) and synthetic N fertilizers (230 and 213 kt N y−1, 83 and 77 kg N ha−1 y−1); in contrast, Sweden used more synthetic (162 kt N y−1, 54 kg N ha−1 y−1) than organic N (108 kt N y−1, 36 kg N ha−1 y−1), with subnational variation in manure use as determined by livestock population. Livestock feed N intake was twice as large in Denmark (384 kt N y−1) as in Sweden (176 kt N y−1), reflecting Denmark’s larger livestock population. Denmark’s national crop NUE was lower (0.51) than Sweden’s (0.72), likely due to a lower proportion of grass–clover leys, higher N input rates, and more intensive production systems. However, considerable subnational variation existed in both countries. The livestock NUE was 0.29 in Denmark and 0.25 in Sweden; these differences are mainly due to a higher proportion of ruminants in Sweden with lower N feed use efficiency than pigs. Sweden emitted less N2O and NH3 per unit area (∼56% for both gases) and in total (∼52% for both gases) than Denmark due to lower use of N inputs and less intensive farming systems. West Denmark and South Sweden were identified as emission hotspots. Our research provides essential information at subnational and national scales to improve N management and reduce gaseous N pollution, supporting the transition towards more sustainable agroecosystems in Denmark and Sweden.

Drosophila melanogaster is a popular research model organism thanks to its' powerful genetic tools that allow spatial and temporal control of gene expression. The inducible GeneSwitch Gal4 system (GS) system is a modified version of the classic UAS/GAL4 system which allows inducible regulation of gene expression and eliminates background effects. It is widely acknowledged that the GS system is leaky, with low level expression of UAS transgenes in absence of the inducer RU-486 (the progesterone analog that activates the modified GAL4 protein). However, in the course of our experiments, we have observed that the extent of this leak depends on the nature of the transgene being expressed. In the absence of RU-486, when strong drivers are used to express protein coding transgenes, leaky expression is low or negligible, however expression of RNA interference (RNAi) transgenes results in complete depletion of protein levels. The majority of published studies, using the GS system and RNAi transgenes validate knock-down efficiency by comparing target gene mRNA levels between induced and non-induced groups. Here, we demonstrate that this approach is lacking and that both additional control groups and further validation is required at the protein level. Unfortunately, this experimental limitation of the GS system eliminates \"the background advantage\", but does offer the possibility of performing more complex experiments (e.g. studying depletion and overexpression of different proteins in the same genetic background). The limitations and new possible applications of the GS system are discussed in detail.

IntroductionThe sunflower broomrape ( Orobanche cumana Wallr.) gene pools of the Guadalquivir Valley and Cuenca province in Spain had predominantly race-F virulence. A new race G was observed recently in the Guadalquivir Valley potentially due to the genetic recombination of the avirulence genes of both gene pools.MethodsIn this research, we have studied populations with atypical virulence from Cuenca. These populations parasitize on DEB2 sunflower line, resistant to all race-G populations evaluated. Ten populations collected in Cuenca province were evaluated with sunflower differential lines and genotyped with 67 SNP markers.ResultsAlthough genetic recombination with individuals of the Guadalquivir Valley gene pool has been observed in most populations, recombination of avirulence genes was discarded as the cause of the new virulence because the population with the highest degree of attack on DEB2 showed no introgression from an external gene pool. Accordingly, a point mutation is proposed as the putative cause of the new virulence.DiscussionThe present study provided a detailed characterization of each population, including the accurate classification of the individuals belonging to each of the classical Spanish gene pools, F1 hybrids, and those that evolved from hybridization between both gene pools. This information is essential to understand how sunflower broomrape populations are evolving in Spain, which in turn may be helpful to understand the dynamics of sunflower broomrape populations in other areas of the world and use this information to develop durable strategies for resistance breeding.

The second-order election (SOE) model as originally formulated by Reif and Schmitt (1980) suggests that, relative to the preceding first-order election result, turnout is lower in SOEs, government and big parties lose, and small and ideologically extreme parties win. These regularities are not static but dynamic and related to the first-order electoral cycle. These predictions of the SOE model have often been tested using aggregate data. The fact that they are based on individual-level hypotheses has received less attention. The main aim of this article is to restate the micro-level hypotheses for the SOE model and run a rigorous test for the 2004 and 2014 European elections. Using data from the European Election Studies voter surveys, our analysis reveals signs of sincere, but also strategic abstentions in European Parliament elections. Both strategic and sincere motivations are also leading to SOE defection. It all happens at once.

Worldwide, Mediterranean cropping systems face the complex challenge of producing enough high-quality food while preserving the quantity and quality of scarce water for people and agriculture in the context of climate change. While good management of nitrogen (N) is paramount to achieving this objective, the efficient strategies developed for temperate systems are often not adapted to the specificities of Mediterranean systems. In this work, we combine original data with a thorough literature review to highlight the most relevant drivers of N dynamics in these semi-arid systems. To do so, we provide an analysis at nested scales combining a bottom-up approach from the field scale, with a top-down approach considering the agro-food system where cropping systems are inserted. We analyze the structural changes in the agro-food systems affecting total N entering the territory, the contrasting response of yields to N availability under rainfed and irrigated conditions in a precipitation gradient, the interaction between N management and climate change adaptation, the main drivers affecting the release of Nr compounds (nitrate, ammonia, nitric oxide and nitrous oxide) compared with temperate systems and finally, the behavior of N once exported to highly regulated river networks. We conclude that sustainable N management in Mediterranean cropping systems requires the specific adaptation of practices to particular local agro-environmental characteristics with special emphasis on water availability for rainfed and irrigated systems. This approach should also include a systemic analysis of N input into the territory that is driven by the configuration of the agro-food system.

Cellular homoeostatic pathways such as macroautophagy (hereinafter autophagy) are regulated by basic mechanisms that are conserved throughout the eukaryotic kingdom. However, it remains poorly understood how these mechanisms further evolved in higher organisms. Here we describe a modification in the autophagy pathway in vertebrates, which promotes its activity in response to oxidative stress. We have identified two oxidation-sensitive cysteine residues in a prototypic autophagy receptor SQSTM1/p62, which allow activation of pro-survival autophagy in stress conditions. The Drosophila p62 homologue, Ref(2)P, lacks these oxidation-sensitive cysteine residues and their introduction into the protein increases protein turnover and stress resistance of flies, whereas perturbation of p62 oxidation in humans may result in age-related pathology. We propose that the redox-sensitivity of p62 may have evolved in vertebrates as a mechanism that allows activation of autophagy in response to oxidative stress to maintain cellular homoeostasis and increase cell survival. The cellular mechanisms underlying autophagy are conserved; however it is unclear how they evolved in higher organisms. Here the authors identify two oxidation-sensitive cysteine residues in the autophagy receptor SQSTM1/p62 in vertebrates which allow activation of pro-survival autophagy in stress conditions.

La Ley 21.563 ha modificado la Ley 20.720, relativa al procedimiento concursal, en diferentes aspectos, siendo el más interesante para este trabajo el que afecta al régimen de los saldos insolutos, cuya regulación entró en vigor el 11 de agosto del 2023. Por su parte, la entrada en vigor en España del texto refundido de la Ley Concursal el 1 de septiembre de 2020 ha supuesto una importante revisión de la exoneración del pasivo insatisfecho. Este trabajo pretende realizar una comparativa, mediante el estudio de la regulación de ambas normativas concursales, del tratamiento que los legisladores nacionales realizan al mecanismo de la segunda oportunidad, con el propósito de realizar propuestas para la posible mejora del tratamiento del régimen de los saldos insolutos.