Problems and directions of improving the insurance of agricultural crops and plantings of perennial plantations against the risks of natural emergencies

The simultaneous existence of two interrelated forms of state aid - subsidizing classical agricultural insurance and compensation for damage to affected regions from the federal budget - requires understanding the basic principles of interaction and assessing the mutual impact of these areas of support. The purpose of the study is to identify the problems of the development of crop insurance and planting of perennial crops against the risks of natural emergencies. The research was carried out in the context of insurance statistics of Russian regions using generally accepted methods. The conducted research revealed the negative dynamics of a significant decrease in the volume of crop insurance of agricultural crops in 2016-2020. Regional authorities in the conditions of a single subsidy are not interested in the independent distribution of funds allocated by the state for the development of agricultural insurance. This is due to the fact that when the agricultural sector has significant losses as a result of the impact of natural disasters, the practice of introducing an emergency regime by the region is widespread. Which gives grounds for receiving compensation for half of the amount of damage directly from the federal budget. This determines the main problem of the development of classical agricultural insurance - the lack of expediency and material interest in this mechanism of protection of property interests both on the part of producers of agricultural products and on the part of regional authorities. In this regard, a brief justification was given for the feasibility of transforming the mechanism for providing direct assistance to the regions from the federal budget into a separate area of crop risk insurance in case of a natural emergency. This line of support will complete the classic multi-risk agricultural insurance.

Divide-and-Conquer Federated Learning Under Data Heterogeneity

Federated Learning allows training of data stored in distributed devices without the need for centralizing training-data, thereby maintaining data-privacy. Addressing the ability to handle data heterogeneity (non-identical and independent distribution or non-IID) is a key enabler for the wider deployment of Federated Learning. In this paper, we propose a novel Divide-andConquer training methodology that enables the use of the popular FedAvg aggregation algorithm by over-coming the acknowledged FedAvg limitations in non-IID environments. We propose a novel use of Cosine-distance based Weight Divergence metric to determine the exact point where a Deep Learning network can be divided into class-agnostic initial layers and class-specific deep layers for performing a Divide and Conquer training. We show that the methodology achieves trained-model accuracy at-par with (and in certain cases exceeding) the numbers achieved by state-of-the-art algorithms like FedProx, FedMA, etc. Also, we show that this methodology leads to compute and/or bandwidth optimizations under certain documented conditions.

STATE OF CENOPOPULATIONS OF SOME RARE SPECIES OF PLANTS OF SAMARA REGION iN THE TERRITORY OF NATURE MONUMENT OF REGIONAL IMPORTANCE "SERNOVODNY SHIKHAN"

The condition of cenopopulations of four rare species ( Artemisia salsoloides , Hedysarum grandiflorum , Oxytropis floribunda , Stipa korshinskyi ) in the petrophytic steppes of the High Zavolzh'yе (High Transvolga, Samara region) was studied on the basis of the regularities of their spatial and ontogenetic structure. The data were processed using the software package R, which allows us to perform calculations of spatial statistics. Thus, stationarity, isotropy, and intensity of plant objects location were evaluated. The mosaic distribution of individuals of all species recorded within the boundaries of the study area was characterized. Using local density maps, the sparseness of individuals on the outskirts of the model site, caused by cattle grazing, was revealed. The behavior of the Ripley's function showed an independent distribution of species as a consequence of weak interspecies competition. The spatial pattern and age spectrum of the studied rare protected dominant species were analyzed. The absence of seedlings and senile plants due to exposure to unfavorable factors of exogenous nature was noted. The im-v and g states were the most numerous. In accordance with the behavior of the K(r) function, the random placement of the four predominant species was revealed, indicating their optimal location within the study area. Patterns of mutual placement of pregenerative and generative individuals of Artemisia salsoloides, Hedysarum grandiflorum, and Oxytropis floribunda were studied. Calculation of the Ripley's cross-function showed that individuals of different age groups of the described species are located independently from each other, demonstrating the absence of intraspecific competition. Undoubtedly, the spatial distribution of individuals of the plant species composing the phytocenosis is influenced both by grazing and by the features of the soil cover of the site, which is manifested by significant elimination of plants at the initial stages of ontogenesis. Nevertheless, the cenopopulations of rare species are stable, mature and promising, and the individuals are distributed in an optimal way that minimizes energy costs.

On-demand biomanufacturing of protective conjugate vaccines

Conjugate vaccines are among the most effective methods for preventing bacterial infections. However, existing manufacturing approaches limit access to conjugate vaccines due to centralized production and cold chain distribution requirements. To address these limitations, we developed a modular technology for in vitro conjugate vaccine expression (iVAX) in portable, freeze-dried lysates from detoxified, nonpathogenic Escherichia coli. Upon rehydration, iVAX reactions synthesize clinically relevant doses of conjugate vaccines against diverse bacterial pathogens in 1 hour. We show that iVAX-synthesized vaccines against Francisella tularensis subsp. tularensis (type A) strain Schu S4 protected mice from lethal intranasal F. tularensis challenge. The iVAX platform promises to accelerate development of new conjugate vaccines with increased access through refrigeration-independent distribution and portable production.

A model for distribution systems performance evaluation

In today's business environment, as the competitiveness becomes increasingly fierce, improving the business performance becomes more important. A major factor affecting business operations is distribution systems performance control. In a distribution network, if an independent point well manages its performance, the efficiency of the entire enterprise will improve. In order to do this, it is essential to evaluate the effectiveness of each distribution point. However, at present, this evaluation is usually based on experience or only a few indicators such as revenue or profit. The effectiveness of a distribution point should be simultaneously considered for sales with resources used such as costs, manpower or business market characteristics such as the number of similar stores within 4 kilometers radius... Therefore, DEA (Data Envelopment Analysis) method is proposed and used to evaluate the performance of independent distribution points and analyze the factors that affect efficiency by the Tobit regression model. This model was tested for a food distribution system in Ho Chi Minh City to prove its feasibility and usefulness.

Decision-Based Marginal Diffusion for Salt-and-Pepper Noise Removal in Color Images

Salt-and-pepper noise suppression for vector-valued images usually employs vector median filtering, total variation L1 model, diffusion methods and variants. These approaches, however, often introduce excessive smoothing and can result in extensive visual feature blurring and are suitable only for images with low intensity noise. In this paper, a new method, as an important preprocessing step in cyber-physical systems, is presented to suppress salt-and-pepper noise that can overcomes this limitation. This method first detects the corrupted pixels and then restores them using channel-wise anisotropic diffusion. The means is twofold. On the one hand, the marginal approach is used to perform noise suppression separately in each channel because the contaminative pixel components are of independent distribution. On the other hand, a decision-based anisotropic diffusion method is applied to each channel to restores them. The anisotropic diffusion is an energy-dissipating process with time, and dependent on geometric analysis of shape of the energy surface. Simulation results indicate that the proposed method for impulsive noise removal achieves the state-of-the-arts results.

Visual Relationship Detection with Low Rank Non-Negative Tensor Decomposition

We address the problem of Visual Relationship Detection (VRD) which aims to describe the relationships between pairs of objects in the form of triplets of (subject, predicate, object). We observe that given a pair of bounding box proposals, objects often participate in multiple relations implying the distribution of triplets is multimodal. We leverage the strong correlations within triplets to learn the joint distribution of triplet variables conditioned on the image and the bounding box proposals, doing away with the hitherto used independent distribution of triplets. To make learning the triplet joint distribution feasible, we introduce a novel technique of learning conditional triplet distributions in the form of their normalized low rank non-negative tensor decompositions. Normalized tensor decompositions take form of mixture distributions of discrete variables and thus are able to capture multimodality. This allows us to efficiently learn higher order discrete multimodal distributions and at the same time keep the parameter size manageable. We further model the probability of selecting an object proposal pair and include a relation triplet prior in our model. We show that each part of the model improves performance and the combination outperforms state-of-the-art score on the Visual Genome (VG) and Visual Relationship Detection (VRD) datasets.

Neutrino masses in an $$SU(4)\otimes U(1)$$-electroweak model with a scalar decuplet

A neutrino mass model is suggested within an $$SU(4)\otimes U(1)$$SU(4)⊗U(1)-electroweak theory. The smallness of neutrino masses can be guaranteed by a seesaw mechanism realized through Yukawa couplings to a scalar SU(4)-decuplet. In this scheme the light active neutrinos are accompanied by heavy neutrinos, which may have masses at different scales, including those within eV–MeV scales investigated quite intensively in both particle physics and astrophysics/cosmology. The flavour neutrinos are superpositions of light neutrinos and a small fraction of heavy neutrinos with the mixing to be determined by the model’s parameters (Yukawa coupling coefficients or symmetry breaking scales). The distribution shape of the Yukawa couplings can be visualized via a model-independent distribution of the neutrino mass matrix elements derived by using the current experimental data. The absolute values of these Yukawa couplings are able to be determined if the symmetry breaking scales are known, and vice versa. With reference to several current and near future experiments, detectable bounds of these heavy neutrinos at different mass scales are discussed and estimated.