Search

This paper combines the Copula-CoVaR approach with the ARMA-GARCH-skewed Student-t model to investigate the tail dependence structure and extreme risk spillover effects between the international agricultural futures and spot markets, taking four main agricultural commodities, namely soybean, maize, wheat, and rice as examples. The empirical results indicate that the tail dependence structures for the four futures-spot pairs are quite different, and each of them exhibits a certain degree of asymmetry. In addition, the futures market for each agricultural commodity has significant and robust extreme downside and upside risk spillover effects on the spot market, and the downside risk spillover effects for both soybeans and maize are significantly stronger than their corresponding upside risk spillover effects, while there is no significant strength difference between the two risk spillover effects for wheat, and rice. This study provides a theoretical basis for strengthening global food cooperation and maintaining global food security, and has practical significance for investors to use agricultural commodities for risk management and portfolio optimization., Comment: 37 pages, 7 figures

Energy security and energy trade are the cornerstones of global economic and social development. The structural robustness of the international oil trade network (iOTN) plays an important role in the global economy. We integrate the machine learning optimization algorithm, game theory, and utility theory for learning an oil trade decision-making model which contains the benefit endowment and cost endowment of economies in international oil trades. We have reconstructed the network degree, clustering coefficient, and closeness of the iOTN well to verify the effectiveness of the model. In the end, policy simulations based on game theory and agent-based model are carried out in a more realistic environment. We find that the export-oriented economies are more vulnerable to be affected than import-oriented economies after receiving external shocks. Moreover, the impact of the increase and decrease of trade friction costs on the international oil trade is asymmetrical and there are significant differences between international organizations., Comment: 14 pages, 5 figures

Pesticides are a kind of agricultural input, whose use can greatly reduce yield loss, regulate plant growth, effectively liberate agricultural productivity, and improve food security. The availability of pesticides in economies all over the world is ensured by pesticide redistribution through international trade and economies play different roles in this process. In this work, we measure and rank the importance of economies using nine node metrics in an evolutionary way. It is found that the clustering coefficient is correlated negatively with the other eight node metrics, while the other eight node metrics are positively correlated with each other and can be grouped into three communities (betweenness; in-degree, PageRank, authority, and in-closeness; out-degree, hub, and out-closeness). We further investigate the structural robustness of the international pesticide trade networks proxied by the giant component size under three types of shocks to economies (node removal in descending order, randomly, and in ascending order). The results show that, except for the clustering coefficient, the international pesticide trade networks are relatively robust under shocks to economies in ascending orders and randomly, but fragile under shocks to economies in descending order. In contrast, removing nodes with the clustering coefficient in ascending and descending orders gives similar robustness curves. Moreover, the structural robustness related to the giant component size evolves over time and exhibits an inverse U-shaped pattern., Comment: 17 pages, 14 figures

Ensuring food supply stability is key to food security for economies, and food imports become increasingly important to safeguard food supplies in economies with inadequate food production. Food import shocks have significant impacts on targeted economies. Using import trade data of four staple crops (maize, rice, soybean, and wheat) from 1986 to 2018, this paper identifies food import trade shocks that occurred to economies during the period of 1995--2018. We compare the temporal evolution and spatial distribution of import shocks occurring to different crops and analyze the shock intensity and shock recovery in various continents based on locally weighted polynomial regression and Cook's distance. The results reveal higher frequencies during the 2007/2008 food crisis and relatively higher shock frequencies in North America, Africa, and Asia. Meanwhile, there are regional differences in shock recovery, with the majority of shocks in Asia recovering in the short term. We also find that high import diversity and a low import dependency ratio buffer economies against import shocks, resulting in a low shock rate and a high recovery rate. These results contribute to our understanding of the external supply risks of food, placing emphasis on accessibility issues in food security.

International food trade is a growing complement to gaps in domestic food supply and demand, but it is vulnerable to disruptions due to some unforeseen shocks. This paper assembles the international crop trade networks using maize, rice, soybean, and wheat trade data sets from 1986 to 2020. We assess the importance of economies using multidimensional node importance metrics. We analyze the correlation structure of different node important metrics based on the random matrix theory and incorporate 20 metrics into a single metric. We find that some metrics have many similarities and dissimilarities, especially for metrics based on the same trade flow directions. We also find that European economies have a significant impact on the iCTNs. Additionally, economies with poor crop production play a major role in import trade, whereas economies with higher food production or smaller populations are crucial to export trade. Our findings have practical implications for identifying key economies in the international crop trade networks, preventing severe damage to the food trade system caused by trade disruptions in some economies, maintaining the stability of the food supply, and ensuring food security.

Pesticides are important agricultural inputs to increase agricultural productivity and improve food security. The availability of pesticides is partially achieved through international trade. However, economies involved in the international trade of pesticides are impacted by internal and external shocks from time to time, which influence the redistribution efficiency of pesticides all over the world. In this work, we adopt simulations to quantify the efficiency and robustness of the international pesticide trade networks under shocks to economies. Shocks are simulated based on nine node metrics, and three strategies are utilized based on descending, random, and ascending node removal. It is found that the efficiency and robustness of the international trade networks of pesticides increased for all the node metrics except the clustering coefficient. Moreover, the international pesticide trade networks are more fragile when import-oriented economies are affected by shocks., Comment: 12 pages, 5 figures

Current class-incremental learning research mainly focuses on single-label classification tasks while multi-label class-incremental learning (MLCIL) with more practical application scenarios is rarely studied. Although there have been many anti-forgetting methods to solve the problem of catastrophic forgetting in class-incremental learning, these methods have difficulty in solving the MLCIL problem due to label absence and information dilution. In this paper, we propose a knowledge restore and transfer (KRT) framework for MLCIL, which includes a dynamic pseudo-label (DPL) module to restore the old class knowledge and an incremental cross-attention(ICA) module to save session-specific knowledge and transfer old class knowledge to the new model sufficiently. Besides, we propose a token loss to jointly optimize the incremental cross-attention module. Experimental results on MS-COCO and PASCAL VOC datasets demonstrate the effectiveness of our method for improving recognition performance and mitigating forgetting on multi-label class-incremental learning tasks.

Deblurring can not only provide visually more pleasant pictures and make photography more convenient, but also can improve the performance of objection detection as well as tracking. However, removing dynamic scene blur from images is a non-trivial task as it is difficult to model the non-uniform blur mathematically. Several methods first use single or multiple images to estimate optical flow (which is treated as an approximation of blur kernels) and then adopt non-blind deblurring algorithms to reconstruct the sharp images. However, these methods cannot be trained in an end-to-end manner and are usually computationally expensive. In this paper, we explore optical flow to remove dynamic scene blur by using the multi-scale spatially variant recurrent neural network (RNN). We utilize FlowNets to estimate optical flow from two consecutive images in different scales. The estimated optical flow provides the RNN weights in different scales so that the weights can better help RNNs to remove blur in the feature spaces. Finally, we develop a convolutional neural network (CNN) to restore the sharp images from the deblurred features. Both quantitative and qualitative evaluations on the benchmark datasets demonstrate that the proposed method performs favorably against state-of-the-art algorithms in terms of accuracy, speed, and model size., Comment: accepted by tcsvt

Machine-to-Machine (M2M) communication is crucial in developing Internet of Things (IoT). As it is well known that cellular networks have been considered as the primary infrastructure for M2M communications, there are several key issues to be addressed in order to deploy M2M communications over cellular networks. Notably, the rapid growth of M2M traffic dramatically increases energy consumption, as well as degrades the performance of existing Human-to-Human (H2H) traffic. Sustainable operation technology and resource management are efficacious ways for solving these issues. In this paper, we investigate a resource management problem in cellular networks with H2H/M2M coexistence. First, considering the energy-constrained nature of machine type communication devices (MTCDs), we propose a novel network model enabled by simultaneous wireless information and power transfer (SWIPT), which empowers MTCDs with the ability to simultaneously perform energy harvesting (EH) and information decoding. Given the diverse characteristics of IoT devices, we subdivide MTCDs into critical and tolerable types, further formulating the resource management problem as an energy efficiency (EE) maximization problem under divers Quality-of-Service (QoS) constraints. Then, we develop a multi-agent deep reinforcement learning (DRL) based scheme to solve this problem. It provides optimal spectrum, transmit power and power splitting (PS) ratio allocation policies, along with efficient model training under designed behaviour-tracking based state space and common reward function. Finally, we verify that with a reasonable training mechanism, multiple M2M agents successfully work cooperatively in a distributed way, resulting in network performance that outperforms other intelligence approaches in terms of convergence speed and meeting the EE and QoS requirements.

We propose brane web configurations for $D$-type and $E$-type $\mathcal{N}=(1,0)$ little string theories based on a trivalent or quadrivalent gluing of 5-brane web diagrams. Tri-/quadri-valent gluing is a powerful way of computing 5d/6d partition functions for supersymmetric gauge theories based on the topological vertex. We generalize the gluing techniques to little string theories by introducing a new compact direction and compute their supersymmetric partition functions on Omega-deformed $\mathbb{R}^4\times T^2$. As concrete examples, we consider little string theories arising from Type IIB NS5-branes probing $D_4$ or $D_5$ singularity. Their effective gauge theory descriptions as the affine $D_4$ or $D_5$ quiver gauge theory can be realized with quadrivalent or trivalent gluing, respectively. Based on these gluings of 5-brane webs, we compute their refined partition functions and compare them with the known results. We extend the computation of the partition function to little string theory engineered from IIB NS5-branes probing $E_6$ singularity based on a trivalent gluing. We also discuss the generalization to higher rank cases and the symmetries of the partition functions., Comment: v2: 97 pages, 32 figures, typos corrected and clarifications added

Only when understanding hackers' tactics, can we thwart their attacks. With this spirit, this paper studies how hackers can effectively launch the so-called 'targeted node attacks', in which iterative attacks are staged on a network, and in each iteration the most important node is removed. In the existing attacks for weighted networks, the node importance is typically measured by the centralities related to shortest-path lengths, and the attack effectiveness is also measured mostly by length-related metrics. However, this paper argues that flows can better reflect network functioning than shortest-path lengths for those networks with carrying traffic as the main functionality. Thus, this paper proposes metrics based on flows for measuring the node importance and the attack effectiveness, respectively. Our node importance metrics include three flow-based centralities (flow betweenness, current-flow betweenness and current-flow closeness), which have not been proposed for use in the attacks on weighted networks yet. Our attack effectiveness metric is a new one proposed by us based on average network flow. Extensive experiments on both artificial and real-world networks show that the attack methods with our three suggested centralities are more effective than the existing attack methods when evaluated under our proposed attack effectiveness metric.

Weight and activation binarization can efficiently compress deep neural networks and accelerate model inference, but cause severe accuracy degradation. Existing optimization methods for binary neural networks (BNNs) focus on fitting full-precision networks to reduce quantization errors, and suffer from the trade-off between accuracy and computational complexity. In contrast, considering the limited learning ability and information loss caused by the limited representational capability of BNNs, we propose IR$^2$Net to stimulate the potential of BNNs and improve the network accuracy by restricting the input information and recovering the feature information, including: 1) information restriction: for a BNN, by evaluating the learning ability on the input information, discarding some of the information it cannot focus on, and limiting the amount of input information to match its learning ability; 2) information recovery: due to the information loss in forward propagation, the output feature information of the network is not enough to support accurate classification. By selecting some shallow feature maps with richer information, and fusing them with the final feature maps to recover the feature information. In addition, the computational cost is reduced by streamlining the information recovery method to strike a better trade-off between accuracy and efficiency. Experimental results demonstrate that our approach still achieves comparable accuracy even with $ \sim $10x floating-point operations (FLOPs) reduction for ResNet-18. The models and code are available at https://github.com/pingxue-hfut/IR2Net., Comment: 11 pages, 8 figures