Search

The decay of sunspot plays a key role in magnetic flux transportation in solar active regions (ARs). To better understand the physical mechanism of the entire decay process of a sunspot, an {\alpha}-configuration sunspot in AR NOAA 12411 was studied. Based on the continuum intensity images and vector magnetic field data with stray light correction from Solar Dynamics Observatory/Helioseismic and Magnetic Imager, the area, vector magnetic field and magnetic flux in the umbra and penumbra are calculated with time, respectively. Our main results are as follows: (1) The decay curves of the sunspot area in its umbra, penumbra, and whole sunspot take the appearance of Gaussian profiles. The area decay rates of the umbra, penumbra and whole sunspot are -1.56 MSH/day, -12.61 MSH/day and -14.04 MSH/day, respectively; (2) With the decay of the sunspot, the total magnetic field strength and the vertical component of the penumbra increase, and the magnetic field of the penumbra becomes more vertical. Meanwhile, the total magnetic field strength and vertical magnetic field strength for the umbra decrease, and the inclination angle changes slightly with an average value of about 20{\deg}; (3) The magnetic flux decay curves of the sunspot in its umbra, penumbra, and whole sunspot exhibit quadratic patterns, their magnetic flux decay rates of the umbra, penumbra and whole sunspot are -9.84 * 10^19 Mx/day, -1.59 * 10^20 Mx/day and -2.60 * 10^20 Mx/day , respectively. The observation suggests that the penumbra may be transformed into the umbra, resulting in the increase of the average vertical magnetic field strength and the reduction of the inclination angle in the penumbra during the decay of the sunspot.

The relationship between the decay of sunspots and moving magnetic features (MMFs) plays an important role in understanding the evolution of active regions. We present observations of two adjacent sunspots, the gap between them, and a lot of MMFs propagating from the gap and the sunspots' outer edges in NOAA Active Region 13023. The MMFs are divided into two types based on their magnetic field inclination angle: vertical (0{\deg}<{\gamma}<45{\deg}) and horizontal (45{\deg}<{\gamma}<90{\deg}) MMFs (V-MMFs and H-MMFs, respectively). The main results are as follows: (1) the mean magnetic flux decay rates of the two sunspots are -1.7*10^20 and -1.4*10^20 Mx/day; (2) the magnetic flux generation rate of all MMFs is calculated to be -1.9 *10^21 Mx/day, which is on average 5.6 times higher than the total magnetic flux loss rate of the sunspots; (3) the magnetic flux of V-MMFs (including a pore separated from the sunspots) is 1.4 times larger than the total lost magnetic flux of the two sunspots, and in a later stage when the pore has passed through the reference ellipse, the magnetic flux generation rate of the V-MMFs is almost the same as the magnetic flux loss rate of the sunspots; and (4) within the gap, the magnetic flux of V-MMFs is one third of the total magnetic flux. Few V-MMFs stream out from the sunspots at the nongap region. All observations suggest that MMFs with vertical magnetic fields are closely related to the disintegration of the sunspot, and most of the MMFs from the gap may originate directly from the sunspot umbra.

Distributional reinforcement learning (DRL) has achieved empirical success in various domains. One of the core tasks in the field of DRL is distributional policy evaluation, which involves estimating the return distribution $\eta^\pi$ for a given policy $\pi$. The distributional temporal difference (TD) algorithm has been accordingly proposed, which is an extension of the temporal difference algorithm in the classic RL literature. In the tabular case, \citet{rowland2018analysis} and \citet{rowland2023analysis} proved the asymptotic convergence of two instances of distributional TD, namely categorical temporal difference algorithm (CTD) and quantile temporal difference algorithm (QTD), respectively. In this paper, we go a step further and analyze the finite-sample performance of distributional TD. To facilitate theoretical analysis, we propose a non-parametric distributional TD algorithm (NTD). For a $\gamma$-discounted infinite-horizon tabular Markov decision process, we show that for NTD we need $\tilde{O}\left(\frac{1}{\varepsilon^{2p}(1-\gamma)^{2p+1}}\right)$ iterations to achieve an $\varepsilon$-optimal estimator with high probability, when the estimation error is measured by the $p$-Wasserstein distance. This sample complexity bound is minimax optimal (up to logarithmic factors) in the case of the $1$-Wasserstein distance. To achieve this, we establish a novel Freedman's inequality in Hilbert spaces, which would be of independent interest. In addition, we revisit CTD, showing that the same non-asymptotic convergence bounds hold for CTD in the case of the $p$-Wasserstein distance.

We study problems of federated control in Markov Decision Processes. To solve an MDP with large state space, multiple learning agents are introduced to collaboratively learn its optimal policy without communication of locally collected experience. In our settings, these agents have limited capabilities, which means they are restricted within different regions of the overall state space during the training process. In face of the difference among restricted regions, we firstly introduce concepts of leakage probabilities to understand how such heterogeneity affects the learning process, and then propose a novel communication protocol that we call Federated-Q protocol (FedQ), which periodically aggregates agents' knowledge of their restricted regions and accordingly modifies their learning problems for further training. In terms of theoretical analysis, we justify the correctness of FedQ as a communication protocol, then give a general result on sample complexity of derived algorithms FedQ-X with the RL oracle , and finally conduct a thorough study on the sample complexity of FedQ-SynQ. Specifically, FedQ-X has been shown to enjoy linear speedup in terms of sample complexity when workload is uniformly distributed among agents. Moreover, we carry out experiments in various environments to justify the efficiency of our methods.

Xiaoxue Zi,1,&ast; Yang Peng,2,&ast; Yiran Zang,1 Shiying Chen,2 Mengshi Li,3 Kena Yu,1 Xu Liang,1 Peng Jin,1 Deyun Wang,4 Li Shi1 1Department of Otolaryngology, The Second Hospital of Shandong University, Jinan, People’s Republic of China; 2State Key Laboratory of Respiratory Disease, National Clinical Research Center for Respiratory Disease, Guangzhou Institute of Respiratory Health, First Affiliated Hospital of Guangzhou Medical University, Guangzhou Medical University, Guangzhou, People’s Republic of China; 3State Key Laboratory of Respiratory Disease, Department of Respiratory Pathology, Guangzhou Institute of Respiratory Health, National Clinical Research Center for Respiratory Disease, The First Affiliated Hospital of Guangzhou Medical University, Guangzhou, People’s Republic of China; 4Department of Otolaryngology, National University of Singapore, National University Health System, Singapore&ast;These authors contributed equally to this workCorrespondence: Peng Jin, Department of Otolaryngology, The Second Hospital of Shandong University, 247 Beiyuan Avenue, Jinan, Shandong, 250033, People’s Republic of China, Tel +86 531 85875317, Fax +86 531 88962544, Email jinpeng0215@126.comObjective: The mechanisms underlying the antrochoanal polyps (ACPs) remained unclear. We aimed to identify the differentially expressed genes (DEGs) profile, the cilia-related genes expression levels and the morphological characteristics of ciliated cells in patients with ACPs.Methods: We obtained ACPs biopsy samples from 28 patients and uncinate process from 27 healthy controls. Whole-transcriptome RNA sequencing, immunofluorescence staining, quantitative polymerase chain reaction, and scanning electron microscopy were performed.Results: 3739 DEGs were detected between ACPs and controls, and Gene Ontology analysis on these DEGs implicated cilium assembly, cilium motility, cilia component, cilia function, inflammatory response and immune system process were includ

Hao Huang,1,&ast; Peng-Yang Yu,1,&ast; Chen Wei,2 Yang-Wei Li,3 Li-Jie Liang,1 Yu-Zhang Liu,1 Li-Na Liu,1 Bai-Jun Fang,1 Yao-Mei Wang1 1Department of Hematology, The Affiliated Cancer Hospital of Zhengzhou University & Henan Cancer Hospital, Zhengzhou, Henan, People’s Republic of China; 2Department of Internal Medicine, The Affiliated Cancer Hospital of Zhengzhou University & Henan Cancer Hospital, Zhengzhou, Henan, People’s Republic of China; 3Central Laboratory, The Affiliated Cancer Hospital of Zhengzhou University & Henan Cancer Hospital, Zhengzhou, Henan, People’s Republic of China&ast;These authors contributed equally to this workCorrespondence: Yao-Mei Wang; Bai-Jun Fang, Email wangyaomeiwym9@126.com; bjfdr_sea@163.comObjective: To examine the clinical characteristics and anemia-related factors in patients with newly diagnosed multiple myeloma (NDMM), as well as the effect and mechanism of erythroblastic islands (EBIs) and EBI macrophages in NDMM patients with anemia.Methods: We collected and analyzed clinical data to find anemia-related factors. Using flow cytometry, the numbers and ratios of erythroblasts and EBI macrophages were determined. RNA sequencing (RNA-seq) was used to determine the differences of EBI macrophages in NDMM patients with or without anemia.Results: Based on the clinical characteristics of NDMM patients with anemia, MCV, abnormal levels of albumin, osteolytic lesions, and Durie-Salmon (DS) stage are risk factors for anemia. Patients with anemia have fewer erythroblasts, erythroblastic islands (EBIs), and EBI macrophages in their bone marrow than patients without anemia. RNA-seq analysis of EBI macrophages from the bone marrow of patients with and without anemia revealed that macrophages from patients with anemia are impaired and tend to promote the production of interleukin-6, which has been demonstrated to be an essential survival factor of myeloma cells and protects them from apoptosis.Conclusion: In NDMM

Autonomous Driving (AD) systems rely on object detection and tracking algorithms that require processing high volumes of data at high frequency. High-performance graphics processing units (GPUs) have been shown to provide the required computing performance. AD also carries functional safety requirements such as diverse redundancy for critical software tasks like object detection. This implies that software must be executed redundantly (in a single GPU for efficiency reasons), and with some form of diversity so that a single fault does not cause the same error in both redundant executions. Unfortunately, high-performance GPUs lack explicit hardware means for diverse redundancy, and software-based solutions with limited guarantees have only been provided for NVIDIA GPUs. This paper presents a software-only solution to enable diverse redundancy on Intel GPUs achieving, for the first time, strong guarantees on the diversity provided. By smartly tailoring workload geometry and managing workload allocation to execution units with thread-level wrappers, we guarantee that redundant threads use physically diverse execution units, hence meeting diverse redundancy requirements with affordable performance overheads., This work has been partially supported by the Spanish Ministry of Science and Innovation under grant PID2019-107255GB-C21/AEI/ 10.13039/501100011033, and by the project AUTOtech.agil of the German Federal Ministry of Education and Research (support code 01IS22088I), Peer Reviewed, Postprint (author's final draft)

Customer-centric marketing campaigns generate a large portion of e-commerce website traffic for Walmart. As the scale of customer data grows larger, expanding the marketing audience to reach more customers is becoming more critical for e-commerce companies to drive business growth and bring more value to customers. In this paper, we present a scalable and efficient system to expand targeted audience of marketing campaigns, which can handle hundreds of millions of customers. We use a deep learning based embedding model to represent customers and an approximate nearest neighbor search method to quickly find lookalike customers of interest. The model can deal with various business interests by constructing interpretable and meaningful customer similarity metrics. We conduct extensive experiments to demonstrate the great performance of our system and customer embedding model.

We propose a novel generalization of constrained Markov decision processes (CMDPs) that we call the \emph{semi-infinitely constrained Markov decision process} (SICMDP). Particularly, we consider a continuum of constraints instead of a finite number of constraints as in the case of ordinary CMDPs. We also devise two reinforcement learning algorithms for SICMDPs that we call SI-CRL and SI-CPO. SI-CRL is a model-based reinforcement learning algorithm. Given an estimate of the transition model, we first transform the reinforcement learning problem into a linear semi-infinitely programming (LSIP) problem and then use the dual exchange method in the LSIP literature to solve it. SI-CPO is a policy optimization algorithm. Borrowing the ideas from the cooperative stochastic approximation approach, we make alternative updates to the policy parameters to maximize the reward or minimize the cost. To the best of our knowledge, we are the first to apply tools from semi-infinitely programming (SIP) to solve constrained reinforcement learning problems. We present theoretical analysis for SI-CRL and SI-CPO, identifying their iteration complexity and sample complexity. We also conduct extensive numerical examples to illustrate the SICMDP model and demonstrate that our proposed algorithms are able to solve complex sequential decision-making tasks leveraging modern deep reinforcement learning techniques., Comment: Shorter version accepted at NeurIPS 2022

Global climate change and ecological environment destruction have become global problems that cannot be ignored. As the largest developing country, China faces the reality of emission reduction from the international community and the pressure of domestic environmental protection. Based on the policy process of carbon peak and carbon neutrality in China, the allocation and use of carbon peak policy tools in China are analyzed by using the Multiple-Streams Approach, and some insights are proposed, hoping to provide some thinking for related research. The initiation of China’s carbon peak and carbon neutrality policy agenda is inseparable from concerns about Problem Streams, discussions on policy streams, and the promotion of Political Streams, as well as the help of policy entrepreneurs. In the future, in the process of continuously promoting carbon peaking and carbon neutrality policies, China should be problem-oriented, continuously optimize policy guidelines, continue to deepen the political environment and promote comprehensive green transformation and sustainable economic and social development with "Double carbon" policies. TRANSLATE with x English Arabic Hebrew Polish Bulgarian Hindi Portuguese Catalan Hmong Daw Romanian Chinese Simplified Hungarian Russian Chinese Traditional Indonesian Slovak Czech Italian Slovenian Danish Japanese Spanish Dutch Klingon Swedish English Korean Thai Estonian Latvian Turkish Finnish Lithuanian Ukrainian French Malay Urdu German Maltese Vietnamese Greek Norwegian Welsh Haitian Creole Persian TRANSLATE with COPY THE URL BELOW Back EMBED THE SNIPPET BELOW IN YOUR SITE Enable collaborative features and customize widget: Bing Webmaster Portal Back

In the process of responding to public health emergencies, the government should promote effective communication with the public and further unblock the channels of communication between the government and the public. Government press conferences are an important channel for the government to communicate with the outside world and an important tool for the government to respond to public health emergencies. The government should make reasonable use of press conferences to actively respond to social concerns, promote the positive function of public opinion, and improve the ability to respond to sudden public health forgetfulness so as to promote the formation of a synergy of social governance. Based on the theory of political communication, this paper focuses on the practical principles and operational risks of government press conferences under public health events and puts forward governance ideas to standardize the release process, strengthen interaction and cooperation, and improve emergency response capacity, so as to achieve emergency management capabilities of government press conferences under public health emergencies.

Utilizing observations from the New Vacuum Solar Telescope (NVST), Solar Dynamics Observatory (SDO), and Solar Terrestrial Relations Observatory-Ahead (STEREO-A), we investigate the event from two distinct observational perspectives: on the solar disk using NVST and SDO, and on the solar limb using STEREO-A. We employ both a non-linear force-free field model and a potential field model to reconstruct the coronal magnetic field, aiming to understand its magnetic properties. Two precursor jet-like activities were observed before the eruption, displaying an untwisted rotation. The second activity released an estimated twist of over two turns. During these two jet-like activities, Y-shaped brightenings, newly emerging magnetic flux accompanied by magnetic cancellation, and the formation of newly moving fibrils were identified. Combining these observational features, it can be inferred that these two precursor jet-like activities released the magnetic field constraining the filament and were triggered by newly emerging magnetic flux. Before the filament eruption, it was observed that some moving flows had been ejected from the site as the onset of two jet-like activities, indicating the same physical process as two jet-like activities. Extrapolations revealed that the filament laid under the height of the decay index of 1.0 and had strong magnetic field (540 Gauss) and a high twisted number (2.4 turns) before the eruption. An apparent rotational motion was observed during the filament eruption. We deduce that the solar filament, exhibiting an inverted U-shape, is a significantly twisted flux rope. The eruption of the filament was initiated by the release of constraining magnetic fields through continuous magnetic reconnection. This reconnection process was triggered by the emergence of newly magnetic flux., Comment: 18 pages, 12 figures, accepted for publication in Astronomy & Astrophysics

Object detection and tracking are vital and fundamental tasks for autonomous driving, aiming at identifying and locating objects from those predefined categories in a scene. 3D point cloud learning has been attracting more and more attention among all other forms of self-driving data. Currently, there are many deep learning methods for 3D object detection. However, the tasks of object detection and tracking for point clouds still need intensive study due to the unique characteristics of point cloud data. To help get a good grasp of the present situation of this research, this paper shows recent advances in deep learning methods for 3D object detection and tracking., Comment: 12 pages, 8 figures

In this paper, we study distributional reinforcement learning from the perspective of statistical efficiency. We investigate distributional policy evaluation, aiming to estimate the complete distribution of the random return (denoted $\eta^\pi$) attained by a given policy $\pi$. We use the certainty-equivalence method to construct our estimator $\hat\eta^\pi$, given a generative model is available. We show that in this circumstance we need a dataset of size $\widetilde O\left(\frac{|\mathcal{S}||\mathcal{A}|}{\epsilon^{2p}(1-\gamma)^{2p+2}}\right)$ to guarantee a $p$-Wasserstein metric between $\hat\eta^\pi$ and $\eta^\pi$ is less than $\epsilon$ with high probability. This implies the distributional policy evaluation problem can be solved with sample efficiency. Also, we show that under different mild assumptions a dataset of size $\widetilde O\left(\frac{|\mathcal{S}||\mathcal{A}|}{\epsilon^{2}(1-\gamma)^{4}}\right)$ suffices to ensure the Kolmogorov metric and total variation metric between $\hat\eta^\pi$ and $\eta^\pi$ is below $\epsilon$ with high probability. Furthermore, we investigate the asymptotic behavior of $\hat\eta^\pi$. We demonstrate that the ``empirical process'' $\sqrt{n}(\hat\eta^\pi-\eta^\pi)$ converges weakly to a Gaussian process in the space of bounded functionals on Lipschitz function class $\ell^\infty(\mathcal{F}_{W_1})$, also in the space of bounded functionals on indicator function class $\ell^\infty(\mathcal{F}_{\mathrm{KS}})$ and bounded measurable function class $\ell^\infty(\mathcal{F}_{\mathrm{TV}})$ when some mild conditions hold. Our findings give rise to a unified approach to statistical inference of a wide class of statistical functionals of $\eta^\pi$.

This paper introduces the Budding Ensemble Architecture (BEA), a novel reduced ensemble architecture for anchor-based object detection models. Object detection models are crucial in vision-based tasks, particularly in autonomous systems. They should provide precise bounding box detections while also calibrating their predicted confidence scores, leading to higher-quality uncertainty estimates. However, current models may make erroneous decisions due to false positives receiving high scores or true positives being discarded due to low scores. BEA aims to address these issues. The proposed loss functions in BEA improve the confidence score calibration and lower the uncertainty error, which results in a better distinction of true and false positives and, eventually, higher accuracy of the object detection models. Both Base-YOLOv3 and SSD models were enhanced using the BEA method and its proposed loss functions. The BEA on Base-YOLOv3 trained on the KITTI dataset results in a 6% and 3.7% increase in mAP and AP50, respectively. Utilizing a well-balanced uncertainty estimation threshold to discard samples in real-time even leads to a 9.6% higher AP50 than its base model. This is attributed to a 40% increase in the area under the AP50-based retention curve used to measure the quality of calibration of confidence scores. Furthermore, BEA-YOLOV3 trained on KITTI provides superior out-of-distribution detection on Citypersons, BDD100K, and COCO datasets compared to the ensembles and vanilla models of YOLOv3 and Gaussian-YOLOv3., Comment: 14 pages, 5 pages supplementary material. Accepted at BMVC-2023

HPC (High Performance Computing) devices increasingly become the only alternative to deliver the performance needed in safety-critical autonomous systems (e.g., autonomous cars, unmanned planes) due to deploying large and powerful multicores along with accelerators such as GPUs. However, the support that those HPC devices offer to realize safety-critical systems on top is heterogeneous. Safety islands have been devised to be coupled to HPC devices and complement them to meet the safety requirements of an increased set of applications, yet the variety of concepts and realizations is large. This paper presents our own concept of a safety island with two goals in mind: (1) offering a wide set of features to enable the broadest set of safety applications for each HPC device, and (2) being realized with open source components based on RISC-V ISA to ease its use and adoption. In particular, we present our safety island concept, the key features we foresee it should include, and its potential application beyond safety., Comment: White paper

Clarifying the evolutionary processes underlying species diversification and adaptation is a key focus of evolutionary biology. Begonia (Begoniaceae) is one of the most species-rich angiosperm genera with c. 2000 species, most of which are shade-adapted. Here, we present chromosome-scale genome assemblies for four species of Begonia (B. loranthoides, B. masoniana, B. darthvaderiana and B. peltatifolia), and whole genome shotgun data for an additional 74 Begonia representatives to investigate lineage evolution and shade adaptation of the genus. The four genome assemblies range in size from 331.75 Mb (B. peltatifolia) to 799.83 Mb (B. masoniana), and harbor 22 059–23 444 protein-coding genes. Synteny analysis revealed a lineage-specific whole-genome duplication (WGD) that occurred just before the diversification of Begonia. Functional enrichment of gene families retained after WGD highlights the significance of modified carbohydrate metabolism and photosynthesis possibly linked to shade adaptation in the genus, which is further supported by expansions of gene families involved in light perception and harvesting. Phylogenomic reconstructions and genomics studies indicate that genomic introgression has also played a role in the evolution of Begonia. Overall, this study provides valuable genomic resources for Begonia and suggests potential drivers underlying the diversity and adaptive evolution of this mega-diverse clade.

Twisted bilayer graphene (TBG) near the magic twist angle of ∼1.1° exhibits a rich phase diagram. However, the interplay between different phases and their dependence on twist angle is still elusive. Here, we explore the stability of various TBG phases and demonstrate that superconductivity near filling of two electrons per moiré unit cell alongside Fermi surface reconstructions, as well as entropy-driven high-temperature phase transitions and linear-in-T resistance occur over a range of twist angles which extends far beyond those exhibiting correlated insulating phases. In the vicinity of the magic angle, we also find a metallic phase that displays a hysteretic anomalous Hall effect and incipient Chern insulating behaviour. Such a metallic phase can be rationalized in terms of the interplay between interaction-driven deformations of TBG bands leading to Berry curvature redistribution and Fermi surface reconstruction. Our results provide an extensive perspective on the hierarchy of correlated phases in TBG as classified by their robustness against deviations from the magic angle or, equivalently, their electronic interaction requirements.

Over the past few years, large knowledge bases have been constructed to store massive amounts of knowledge. However, these knowledge bases are highly incomplete, for example, over 70% of people in Freebase have no known place of birth. To solve this problem, we propose a query-driven knowledge base completion system with multimodal fusion of unstructured and structured information. To effectively fuse unstructured information from the Web and structured information in knowledge bases to achieve good performance, our system builds multimodal knowledge graphs based on question answering and rule inference. We propose a multimodal path fusion algorithm to rank candidate answers based on different paths in the multimodal knowledge graphs, achieving much better performance than question answering, rule inference and a baseline fusion algorithm. To improve system efficiency, query-driven techniques are utilized to reduce the runtime of our system, providing fast responses to user queries. Extensive experiments have been conducted to demonstrate the effectiveness and efficiency of our system.

Over the past few years, large knowledge bases have been constructed to store massive amounts of knowledge. However, these knowledge bases are highly incomplete. To solve this problem, we propose a web-based question answering system system with multimodal fusion of unstructured and structured information, to fill in missing information for knowledge bases. To utilize unstructured information from the Web for knowledge base completion, we design a web-based question answering system using multimodal features and question templates to extract missing facts, which can achieve good performance with very few questions. To help improve extraction quality, the question answering system employs structured information from knowledge bases, such as entity types and entity-to-entity relatedness.

Object detection neural network models need to perform reliably in highly dynamic and safety-critical environments like automated driving or robotics. Therefore, it is paramount to verify the robustness of the detection under unexpected hardware faults like soft errors that can impact a systems perception module. Standard metrics based on average precision produce model vulnerability estimates at the object level rather than at an image level. As we show in this paper, this does not provide an intuitive or representative indicator of the safety-related impact of silent data corruption caused by bit flips in the underlying memory but can lead to an over- or underestimation of typical fault-induced hazards. With an eye towards safety-related real-time applications, we propose a new metric IVMOD (Image-wise Vulnerability Metric for Object Detection) to quantify vulnerability based on an incorrect image-wise object detection due to false positive (FPs) or false negative (FNs) objects, combined with a severity analysis. The evaluation of several representative object detection models shows that even a single bit flip can lead to a severe silent data corruption event with potentially critical safety implications, with e.g., up to (much greater than) 100 FPs generated, or up to approx. 90% of true positives (TPs) are lost in an image. Furthermore, with a single stuck-at-1 fault, an entire sequence of images can be affected, causing temporally persistent ghost detections that can be mistaken for actual objects (covering up to approx. 83% of the image). Furthermore, actual objects in the scene are continuously missed (up to approx. 64% of TPs are lost). Our work establishes a detailed understanding of the safety-related vulnerability of such critical workloads against hardware faults., Comment: 15 pages, accepted in safecomp22 conference

We study a Federated Reinforcement Learning (FedRL) problem in which $n$ agents collaboratively learn a single policy without sharing the trajectories they collected during agent-environment interaction. We stress the constraint of environment heterogeneity, which means $n$ environments corresponding to these $n$ agents have different state transitions. To obtain a value function or a policy function which optimizes the overall performance in all environments, we propose two federated RL algorithms, \texttt{QAvg} and \texttt{PAvg}. We theoretically prove that these algorithms converge to suboptimal solutions, while such suboptimality depends on how heterogeneous these $n$ environments are. Moreover, we propose a heuristic that achieves personalization by embedding the $n$ environments into $n$ vectors. The personalization heuristic not only improves the training but also allows for better generalization to new environments., Comment: Artificial Intelligence and Statistics 2022

Flat electronic bands, characteristic of magic-angle twisted bilayer graphene (TBG), host a wealth of correlated phenomena. Early theoretical considerations suggested that, at the magic angle, the Dirac velocity vanishes and the entire width of the moiré bands becomes extremely narrow. Yet, this scenario contradicts experimental studies that reveal a finite Dirac velocity as well as bandwidths significantly larger than predicted. Here we use spatially resolved spectroscopy in finite and zero magnetic fields to examine the electronic structure of moiré bands and their intricate connection to correlated phases. By following the relative shifts of Landau levels in finite fields, we detect filling-dependent band flattening, that unexpectedly starts already at ~1.3 degrees, well above the magic angle and hence nominally in the weakly correlated regime. We further show that, as the twist angle is reduced, the moiré bands become maximally flat at progressively lower doping levels. Surprisingly, when the twist angles reach values for which the maximal flattening occurs at approximate filling of −2, +1,+2,+3 electrons per moiré unit cell, the corresponding zero-field correlated phases start to emerge. Our observations are corroborated by calculations that incorporate an interplay between the Coulomb charging energy and exchange interactions; together these effects produce band flattening and hence a significant density-of-states enhancement that facilitates the observed symmetry-breaking cascade transitions. Besides emerging phases pinned to integer fillings, we also experimentally identify a series of pronounced correlation-driven band deformations and soft gaps in a wider doping range around ±2 filling where superconductivity is expected. Our results highlight the role of interaction-driven band-flattening in forming robust correlated phases in TBG.

Spinel oxides are considered as promising low-cost non-precious metal electrocatalysts for oxygen evolution reaction (OER) due to their desirable catalytic activities and fast kinetics. However, as a result of the structural complexity of spinel oxides, systematic and in-depth studies on enhancing the OER performance of spinel oxides remain inadequate. In particular, the construction of active sites regarding the large number of unoccupied octahedral interstices has not yet been explored. Herein, more octahedral sites with high OER activities are constructed on the surface of spinel oxides via a cationic misalignment, which is induced by the defects in the spinel oxide solutions, i.e., MoFe2O4 and CoFe2O4 nanosheets supported on an iron foam (MCFO NS/IF). With increased active sites and modified electronic structure, the state-of-the-art electrocatalyst exhibits the excellent OER catalytic activity with an onset potential of 1.41 V versus RHE and an overpotential of 290 mV to achieve a current density of 500 mA cm−2. Moreover, such an electrocatalyst also demonstrates fast kinetics with the Tafel slope of 38 mV dec−1 and superior durability by maintaining the OER activity at 250 mA cm−2 for 1000 h. © 2021 The Authors. Advanced Science published by Wiley-VCH GmbH.

A magnetic separation–shaking table gravity separation process was used to treat Sc-bearing V–Ti magnetite tailings, and the Sc2O3 content in the Sc concentrate was enhanced from 0.0039 to 0.0061%. The majority of the Sc in the Sc concentrate originated from augite, hornblende, and aluminosilicate. Magnesium chloride and the Sc concentrate were placed in a roasting furnace for chloridising roasting, causing the partial chlorination of Sc to ScCl3. Sc was extracted from the roasted ore using hydrochloric acid leaching. Sc extraction indices demonstrated 95.12% Sc2O3 leaching efficiency and 0.0011% Sc2O3 content in the acid leaching residue, clearly indicating the extraction of Sc. No noticeable peak of Sc was observed in the spectrum of the acid leaching residue. Most of the Sc in the roasted ores was dissolved as Sc3+, and it entered the liquid phase. The main minerals in the acid leaching residue were Mg2SiO4, CaTiO3, and Fe2SiO4. (Authors.), A magnetic separation–shaking table gravity separation process was used to treat Sc-bearing V–Ti magnetite tailings, and the Sc2O3 content in the Sc concentrate was enhanced from 0.0039 to 0.0061%. The majority of the Sc in the Sc concentrate originated from augite, hornblende, and aluminosilicate. Magnesium chloride and the Sc concentrate were placed in a roasting furnace for chloridising roasting, causing the partial chlorination of Sc to ScCl3. Sc was extracted from the roasted ore using hydrochloric acid leaching. Sc extraction indices demonstrated 95.12% Sc2O3 leaching efficiency and 0.0011% Sc2O3 content in the acid leaching residue, clearly indicating the extraction of Sc. No noticeable peak of Sc was observed in the spectrum of the acid leaching residue. Most of the Sc in the roasted ores was dissolved as Sc3+, and it entered the liquid phase. The main minerals in the acid leaching residue were Mg2SiO4, CaTiO3, and Fe2SiO4. (Authors.)

Purpose: Australian Aboriginal and Torres Strait Islander diagnosed with cancer are known to experience poorer survival, with these survival disparities mainly restricted to the first 2 years after diagnosis. With improved accuracy and completeness of identifying Aboriginal and Torres Strait Islander peoples over the whole study period, our goal was to examine whether the survival disparity among Aboriginal and Torres Strait Islander peoples diagnosed with cancer in Queensland has changed over time. Methods: Population-based data from the Queensland Cancer Register between 1998 and 2017 for Queenslanders aged 15 years and over at diagnosis (n = 377,963; 1.6% Aboriginal and/or Torres Strait Islander) were used to determine whether this disparity has reduced over time. Flexible parametric survival models incorporating time-varying coefficients were used to examine the association between Aboriginal and Torres Strait Islander status and cancer-specific survival within 5 years of diagnosis. Results: The adjusted 5-year cancer-specific survival rate for Aboriginal and Torres Strait Islander people diagnosed with cancer increased from 60.5% (95% CI 59.2–61.9%) in 1998–2007 to 65.5% (95% CI 64.3–66.6%) in 2008–2017, with the corresponding estimates for other Queenslanders being 66.6% (95% CI 66.4–66.8%) and 70.1% (95% CI 69.9–70.3%). The survival disparity was significant only for the first 3 years since diagnosis for 1998–2007; however, it was significantly (p ≤ 0.02) elevated for all five time intervals for 2008–2017, with similar average hazard ratios (95% CIs) over the 5-year interval after diagnosis of 1.45 (1.36–1.55) for 1998–2007 and 1.42 (1.34–1.50) for 2008–2017. Conclusion: Although survival has increased over time, the lack of improvement in the disparity in cancer survival experienced by Aboriginal and Torres Strait Islander cancer patients highlights the urgent need to better understand the multifaceted and completed factors that underlie this gap to guide

Graphene moire superlattices have emerged as a platform hosting and abundance of correlated insulating, topological, and superconducting phases. While the origins of strong correlations and non-trivial topology are shown to be directly linked to flat moire bands, the nature and mechanism of superconductivity remain enigmatic. In particular, only alternating twisted stacking geometries of bilayer and trilayer graphene are found to exhibit robust superconductivity manifesting as zero resistance and Fraunhofer interference patterns. Here we demonstrate that magic-angle twisted tri-, quadri-, and pentalayers placed on monolayer tungsten diselenide exhibit flavour polarization and superconductivity. We also observe insulating states in the trilayer and quadrilayer arising at finite electric displacement fields, despite the presence of dispersive bands introduced by additional graphene layers. Moreover, the three multilayer geometries allow us to identify universal features in the family of graphene moire structures arising from the intricate relations between superconducting states, symmetry-breaking transitions, and van Hove singularities. Remarkably, as the number of layers increases, superconductivity emerges over a dramatically enhanced filling-factor range. In particular, in twisted pentalayers, superconductivity extends well beyond the filling of four electrons per moire unit cell, demonstrating the non-trivial role of the additional bands. Our results highlight the importance of the interplay between flat and dispersive bands in extending superconducting regions in graphene moire superlattices and open new frontiers for developing graphene-based superconductors.

Unsupervised domain adaptive person re-identification has received significant attention due to its high practical value. In past years, by following the clustering and finetuning paradigm, researchers propose to utilize the teacher-student framework in their methods to decrease the domain gap between different person re-identification datasets. Inspired by recent teacher-student framework based methods, which try to mimic the human learning process either by making the student directly copy behavior from the teacher or selecting reliable learning materials, we propose to conduct further exploration to imitate the human learning process from different aspects, \textit{i.e.}, adaptively updating learning materials, selectively imitating teacher behaviors, and analyzing learning materials structures. The explored three components, collaborate together to constitute a new method for unsupervised domain adaptive person re-identification, which is called Human Learning Imitation framework. The experimental results on three benchmark datasets demonstrate the efficacy of our proposed method.

Convolutional neural networks (CNNs) have become an established part of numerous safety-critical computer vision applications, including human robot interactions and automated driving. Real-world implementations will need to guarantee their robustness against hardware soft errors corrupting the underlying platform memory. Based on the previously observed efficacy of activation clipping techniques, we build a prototypical safety case for classifier CNNs by demonstrating that range supervision represents a highly reliable fault detector and mitigator with respect to relevant bit flips, adopting an eight-exponent floating point data representation. We further explore novel, non-uniform range restriction methods that effectively suppress the probability of silent data corruptions and uncorrectable errors. As a safety-relevant end-to-end use case, we showcase the benefit of our approach in a vehicle classification scenario, using ResNet-50 and the traffic camera data set MIOVision. The quantitative evidence provided in this work can be leveraged to inspire further and possibly more complex CNN safety arguments., Comment: 8 pages, 7 figures

Endometrial cancer (EC) is the fifth most common cancer in women in developed countries. Clinical practice guidelines recommend patients should be followed-up every 3–6 months after primary treatment of EC. Evidence suggests that 40% to 80% of patients develop symptoms prior to being diagnosed with EC recurrence, however which symptoms are key remains unclear. We previously conducted a comprehensive literature review and developed a questionnaire on patient-reported symptoms associated with EC recurrence. This is a brief communication on a pilot prospective cohort study among 120 Australian patients who completed primary treatment for EC in the past three years. The study showed 47 of the 120 patients (39.2%) self-reported at least one symptom, four of whom (3.3%) were diagnosed with a recurrence. Back or lumbar pain (P = 0.012), vaginal bleeding (P < 0.001), and lethargy, fatigue, exhaustion or tiredness (P = 0.002) were significantly associated with the development of EC recurrence. The checklist will be further validated as part of a randomized controlled clinical trial to confirm the observed relationship between symptoms and the development of EC recurrence.

To construct a robot that can walk as efficiently and steadily as humans or other legged animals, we develop an enhanced elitist-mutated ant colony optimization~(EACO) algorithm with genetic and crossover operators in real-time applications to humanoid robotics or other legged robots. This work presents promoting global search capability and convergence rate of the EACO applied to humanoid robots in real-time by estimating the expected convergence rate using Markov chain. Furthermore, we put a special focus on the EACO algorithm on a wide range of problems, from ACO, real-coded GAs, GAs with neural networks~(NNs), particle swarm optimization~(PSO) to complex robotics systems including gait synthesis, dynamic modeling of parameterizable trajectories and gait optimization of humanoid robotics. The experimental results illustrate the capability of this method to discover the premature convergence probability, tackle successfully inherent stagnation, and promote the convergence rate of the EACO-based humanoid robotics systems and demonstrated the applicability and the effectiveness of our strategy for solving sophisticated optimization tasks. We found reliable and fast walking gaits with a velocity of up to 0.47m/s using the EACO optimization strategy. These findings have significant implications for understanding and tackling inherent stagnation and poor convergence rate of the EACO and provide new insight into the genetic architectures and control optimization of humanoid robotics., Comment: 20 pages, 10 figures, 53 references

One of the ambitions of artificial intelligence is to root artificial intelligence deeply in basic science while developing brain-inspired artificial intelligence platforms that will promote new scientific discoveries. The challenges are essential to push artificial intelligence theory and applied technologies research forward. This paper presents the grand challenges of artificial intelligence research for the next 20 years which include:~(i) to explore the working mechanism of the human brain on the basis of understanding brain science, neuroscience, cognitive science, psychology and data science; (ii) how is the electrical signal transmitted by the human brain? What is the coordination mechanism between brain neural electrical signals and human activities? (iii)~to root brain-computer interface~(BCI) and brain-muscle interface~(BMI) technologies deeply in science on human behaviour; (iv)~making research on knowledge-driven visual commonsense reasoning~(VCR), develop a new inference engine for cognitive network recognition~(CNR); (v)~to develop high-precision, multi-modal intelligent perceptrons; (vi)~investigating intelligent reasoning and fast decision-making systems based on knowledge graph~(KG). We believe that the frontier theory innovation of AI, knowledge-driven modeling methodologies for commonsense reasoning, revolutionary innovation and breakthroughs of the novel algorithms and new technologies in AI, and developing responsible AI should be the main research strategies of AI scientists in the future., Comment: 13 pages; 7 figures; 23 references

Based on the similarity theory, a physical testing model with scale of 1:10 was designed and fabricated. The Wenchuan wave (WC-XZ), Darui artificial wave (DR-XZ) and Kobe wave (K-XZ) were adopted as excitation waves which was XZ directions (horizontal and vertical loading at the same time). A series of shaking table tests were carried out to study the effect of the types, directions and peaks acceleration of the seismic waves on the acceleration response laws of the shallow-buried bias tunnels with small distance. The results show that: (1) in the horizontal direction, the types of the seismic waves have a little effect on the acceleration response, but the law is opposite with Kobe wave being the most serious, followed by Darui artificial wave and Wenchuan wave in the vertical direction. (2) The existing slope has a significant effect on the acceleration response of the right hole which is different from that of the left hole. (3) The acceleration response of the tunnel in the vertical direction is more severe than that of the horizontal direction and the acceleration amplification factors in vertical direction are is generally 1.02-3.94 times that of the horizontal direction. (4) in the seismic design of tunnel, different seismic measures should be adopted in different directions. The results have a significant reference for the anti-seismic design of the tunnel.

Acral melanoma (AM) is an extremely aggressive subtype of melanoma that is prevalent in eastern Asia. AM exhibits high intertumoral and intratumoral heterogeneities with poor prognosis. To associate the genomic heterogeneities with phenotypic traits and efficacy of treatments, a method is needed to recover genomic information from limited samples with high specificity and sensitivity from early stage AM specimens. We performed laser capture microdissection to isolate single micro-tumor nests, containing only dozens of cells, from stained tissue slices and then applied multiple annealing and looping-based amplification cycles, a highly efficient whole-genome amplification method originally developed for single cells, to amplify the whole genome of each tumor nest for sequencing. We were able to accurately profile the landscape of copy number alterations and single nucleotide variations of every single micro-tumor nest and to quantitatively characterize the heterogeneities at different levels, between tumor and nevi, among patients, among different phenotypes within a same tumor, and among adjacent tumor cell clusters with identical phenotypic appearance. We have found that genomic heterogeneity exists extensively and that branched evolution happens in the early stage of AM development. We are able to build the phylogenetic tree among these phenotypically addressable cell clusters. © 2019 The Authors

Transitional metal dichacogenides (TMDs) have stimulated an increasing research and technological attention due to their unique properties, holding great promise for emerging energy storage and conversion applications. However, tailorable and efficient synthesis of TMDs to garner the electrochemical and electrocatalytic performance of thus-derived electrodes has by far remained challenging. Herein we demonstrate a versatile synthetic strategy to in situ grow CoSe 2 @vertically-oriented graphene (VG) hierarchical architecture on carbon fiber cloth (CC) via combined steps of plasma-enhanced chemical vapor deposition and wet chemistry. Such self-supporting and flexible CoSe 2 @VG/CC arrays possess significant implications for pseudocapacitive Na storage and electrocatalytic O 2 evolution (OER). When evaluated as an anode material for sodium-ion hybrid capacitors, full cells comprising a CoSe 2 @VG/CC anode and AC cathode enable a favorable cyclic stability at 0.5 A g −1 for 1800 cycles in the potential range of 0.5-3.3 V, harvesting a high energy and power density of 116 Wh kg −1 and 7298 W kg −1 . In addition, CoSe 2 @VG/CC array also exhibits an excellent OER performance with a low overpotential of 418 mV and Tafel slope of 82 mV dec −1 on a basis of experimental exploration and theoretical simulation.

Yang Peng, Zhiqiang Wang Centre for Chronic Disease, Faculty of Medicine, The University of Queensland, Brisbane, QLD, Australia Background: The American Heart Association (AHA) committee recently set a guideline to define and monitor the cardiovascular health status. This study aimed to estimate cardiovascular health status among Australian adults using the guideline.Methods: We used data from a nationally representative sample of 7499 adults (age ≥18 years) from 2011 to 2012 Australian Health Survey. We applied the modified AHA’s definition to estimate the ideal proportions of the seven metrics and the overall cardiovascular health status.Results: Ideal status was most prevalent for fasting plasma glucose (83.6%) and least observed for dietary pattern (4.8%). The estimated percentage of ideal cardiovascular health was 0.15% in Australian adults. An estimated 0.52% of Australian adults had all four ideal cardiovascular health factors, and 16.38% had all four ideal cardiovascular health behaviors. There exist some age and sex variations for the ideal status of individual metric and the overall cardiovascular health.Conclusion: The percentage of ideal cardiovascular health in Australian adults is extremely low. Public health policies should be implemented to improve the population-wide cardiovascular health status in Australia. Keywords: cardiovascular health, Australian adults, cross-sectional study, risk factors&nbsp

ARID1A, a component of the chromatin remodeling complex SWI/SNF, is an evolutionarily conserved complex that uses the energy of adenosine triphosphate hydrolysis to remodel chromatin structure and functions as a master regulator of gene transcription. Recent genomic studies have revealed that ARID1A is one of the most frequently mutated genes in human cancers. However, therapeutic approaches that selectively target ARID1A-mutant tumors are not yet clinically available. Our previous study showed that ARID1A facilitates chromatin response and cell cycle checkpoint activation after DNA damage. Therefore, an ARID1A deficiency may result in therapeutic vulnerabilities in cell cycle modulators. The goals of our study were to develop a novel screening approach, based on fluorescent ubiquitination-based cell cycle indicators (FUCCI), and to identify chemical agents that can selectively modulate the cell cycle transition in ARID1A-deficient cancer cells. Using this high-throughput assay, we screened 2643 compounds and identified six potential chemical modulators that can selectively modulate the cell cycle in ARID1A-deficient cells; these agents may be useful for developing new therapeutics for ARID1A-mutant tumors. In summary, our study demonstrates that FUCCI cell-based high-content screening is a powerful and effective approach for identifying cell cycle modulators and can be applied to multigenotypic screening for targeted cancer therapeutics.

The development of rational synthetic procedures with desired nuclearity and high selectivity is a critical issue in inorganic chemistry. Here we demonstrate a comprehensive understanding of the template effect induced by metal cations in the formation mechanism of the class of polyoxopalladates ({MPd12L8} nanocube and {MPd15L10} nanostar) by combining computational and experimental techniques. The capture of a Mn+ guest ion by a peripheral palladium(II)-oxo shell leads to a competition between the parent Pd2+ addenda ion and the respective guest metal ion. The present study reveals that (i) the selection of the incorporated guest ion has a thermodynamic control, (ii) the main factors governing the formation of a particular polyanion are the charge and size of the guest cation, (iii) the electrostatic interaction between the cation and the surrounding oxo ligands and (iv) the dehydration ability of the cation. As expected from the number of observed {Mn+Pd12L8} species, trivalent cations M3+ were found to be good templates resulting in several examples of {M3+Pd12L8}, whereas monovalent cations M+ are much less prone to form {M+Pd12L8}. For tetravalent cations the dehydration energies are very large, however, the formation of {M4+Pd12L8} nanocubes is found to be still energetic favourable. Fully consistent with computational predictions, four novel polyoxo-12-palladates were synthesized: the La3+-centered nanocube [LaPd12O8(PhAsO3)8]5− (LaPd12-closed), the La3+-centered “open” nanocube [LaPd12O6(OH)3(PhAsO3)6(OAc)3]3− (LaPd12-open), the Ga3+-centered [GaPd12O8(PhAsO3)8]5− (GaPd12), and the In3+-analogue [InPd12O8(PhAsO3)8]5− (InPd12). All four compounds were fully characterized in the solid state and in solution by a multitude of physicochemical techniques, including 71

In this paper, we present our effort towards comprehensive traffic forecasting for big data applications using external, light-weighted file system monitoring. Our idea is motivated by the key observations that rich traffic demand information already exists in the log and meta-data files of many big data applications, and that such information can be readily extracted through run-time file system monitoring. As the first step, we use Hadoop as a concrete example to explore our methodology and develop a system called HadoopWatch to predict traffic demands of Hadoop applications. We further implement HadoopWatch in a small-scale testbed with 10 physical servers and 30 virtual machines. Our experiments over a series of MapReduce applications demonstrate that HadoopWatch can forecast the traffic demand with almost 100% accuracy and time advance. Furthermore, it makes no modification on the Hadoop framework, and introduces little overhead to the application performance. Finally, to showcase the utility of accurate traffic prediction made by HadoopWatch, we design and implement a simple HadoopWatch-enabled network optimization module into the HadoopWatch controller, and with realistic Hadoop job benchmarks we find that even a simple algorithm can leverage the forecasting results provided by HadoopWatch to significantly improve the Hadoop job completion time by up to 14.72%. © 2015 IEEE.

This paper analyzes influences on the credit standing of Multilateral Development Banks (MDBs), specifically the quality, diversification and single name concentration of their portfolios, and on the market practice known as Preferred Creditor Status (PCS), whereby sovereigns that default on other debt rarely fail to meet their obligations to MDBs. The paper examines how rating agencies assess MDB ratings, focusing in particular on how Standard & Poor's assesses capital as part of its MDB rating process. The results are compared with those implied by an industry-standard, ratings-based Credit Risk Model (CRM). Applying these methodologies to one specific MDB, the Inter-American Development Bank (IDB), the paper shows that S&P's approach is highly conservative in its evaluation of single name concentration risk and makes insufficient allowance for PCS. Calibrating the CRM with risk-neutral distributions, the paper examines the effect of PCS on MDB funding rates and compares model-implied funding spreads with those observed in the market.

UnlabelledARID1A, SWI/SNF chromatin remodeling complex subunit, is a recently identified tumor suppressor that is mutated in a broad spectrum of human cancers. Thus, it is of fundamental clinical importance to understand its molecular functions and determine whether ARID1A deficiency can be exploited therapeutically. In this article, we report a key function of ARID1A in regulating the DNA damage checkpoint. ARID1A is recruited to DNA double-strand breaks (DSB) via its interaction with the upstream DNA damage checkpoint kinase ATR. At the molecular level, ARID1A facilitates efficient processing of DSB to single-strand ends and sustains DNA damage signaling. Importantly, ARID1A deficiency sensitizes cancer cells to PARP inhibitors in vitro and in vivo, providing a potential therapeutic strategy for patients with ARID1A-mutant tumors.SignificanceARID1A has been identified as one of the most frequently mutated genes across human cancers. Our data suggest that clinical utility of PARP inhibitors might be extended beyond patients with BRCA mutations to a larger group of patients with ARID1A-mutant tumors, which may exhibit therapeutic vulnerability to PARP inhibitors.

As the world enters the era of big data, MapReduce-like data-parallel computing frameworks are widely adopted in clouds and data centers. My objective is to build a more efficient underlying networks and improve the performance of such computing system. Specically, I first present my effort towards comprehensive traffic forecasting for big data applications using external, light-weighted file system monitoring. The idea is motivated by the key observations that rich traffic demand information already exists in the log and meta-data files of many big data applications, and that such information can be readily extracted through run-time file system monitoring. As an initial step, we use Hadoop1 as a concrete example to explore our methodology and develop a system called HadoopWatch to predict traffic demand of Hadoop applications. Our experiments over a series of MapReduce applications demonstrate that HadoopWatch can forecast the traffic demand with almost 100% accuracy and time advance. Meanwhile, it makes no modication of the Hadoop framework, and introduces little overhead to the application performance. Second, I also make my attempt to orchestrate the network with the traffic demand information forecasted by HadoopWatch. After studying the traffic pattern widely existed in these computing frameworks, I realize that many flows are grouped to achieve a common barrier. To diminish the average group communication time, I proposed a task-aware flow scheduling and routing scheme. Flow scheduling is based on Shortest Remaining First scheduling paradigm, while task-aware flow routing can guarantee the grouped flows are evenly distributed across the multiple paths. I design and implement a prototype system called ShuffleBoost, which can cooperate with HadoopWatch and improve network efficiency in Hadoop clusters. In the best case I measured, the average group communication time decreased by 13.5%, and the average job completion time drops by 7.8%. 1

This paper presents our effort towards comprehensive traffic forecasting for big data applications using external, light-weighted file system monitoring. Our idea is motivated by the key observations that rich traffic demand information already exists in the log and meta-data files of many big data applications, and that such information can be readily extracted through run-time file system monitoring. As the first step, we use Hadoop1 as a concrete example to explore our methodology and develop a system called HadoopWatch to predict traffic demand of Hadoop applications. We further implement HadoopWatch in our real small-scale testbed with 10 physical servers and 30 virtual machines. Our experiments over a series of MapReduce applications demonstrate that HadoopWatch can forecast the traffic demand with almost 100% accuracy and time advance. Furthermore, it makes no modification of the Hadoop framework, and introduces little overhead to the application performance. © 2014 IEEE.

This paper presents our effort towards comprehensive traffic forecasting for big data applications using external, light-weighted file system monitoring. Our idea is motivated by the key observations that rich traffic demand information already exists in the log and meta-data files of many big data applications, and that such information can be readily extracted through run-time file system monitoring. As the first step, we use Hadoop1 as a concrete example to explore our methodology and develop a system called HadoopWatch to predict traffic demand of Hadoop applications. We further implement HadoopWatch in our real small-scale testbed with 10 physical servers and 30 virtual machines. Our experiments over a series of MapReduce applications demonstrate that HadoopWatch can forecast the traffic demand with almost 100% accuracy and time advance. Furthermore, it makes no modification of the Hadoop framework, and introduces little overhead to the application performance. © 2014 IEEE.

A Cu?carborane-based metal?organic framework (MOF), NU-135, which contains a quasi-spherical para-carborane moiety, has been synthesized and characterized. NU-135 exhibits a pore volume of 1.02 cm3/g and a gravimetric BET surface area of ca. 2600 m2/g, and thus represents the first highly porous carborane-based MOF. As a consequence of the unique geometry of the carborane unit, NU-135 has a very high volumetric BET surface area of ca. 1900 m2/cm3. CH4, CO2, and H2 adsorption isotherms were measured over a broad range of pressures and temperatures and are in good agreement with computational predictions. The methane storage capacity of NU-135 at 35 bar and 298 K is ca. 187 vSTP/v. At 298 K, the pressure required to achieve a methane storage density comparable to that of a compressed natural gas (CNG) tank pressurized to 212 bar, which is a typical storage pressure, is only 65 bar. The methane working capacity (5?65 bar) is 170 vSTP/v. The volumetric hydrogen storage capacity at 55 bar and 77 K is 49 g/L. These properties are comparable to those of current record holders in the area of methane and hydrogen storage. This initial example lays the groundwork for carborane-based materials with high surface areas., The original document contains color images.

The objective of this thesis is to achieve improved Quality of Experience (QoE) for video services with stringent delay requirement, delivered over time-varying error-prone communication channels. A low-delay error-resilient video transmission framework is proposed for point-to-point communication with feedback information. The proposed framework consists of low-delay error-resilient video coding and delay-aware retransmission, where multi-dimensional video adaptation and joint source-channel resource allocation schemes are formulated to improve the channel adaptability. A full-reference video quality metric is developed and integrated into the framework for QoE estimation in the presence of both spatial and temporal quality impairments. Experimental results show that the QoE-driven system delivers significantly improved QoE for a wide variety of video contents and channel conditions., Das Ziel dieser Arbeit ist es, die Nutzerzufriedenheit (QoE) für Video-Dienste mit strikten Verzögerungsanforderungen, die über zeitveränderliche und fehleranfällige Kommunikationskanäle bereitgestellt werden, zu verbessern. Ein fehlerrobustes Videoübertragungssystem mit geringer Verzögerung wird für Punkt-zu-Punkt-Kommunikation mit Feedbackinformation vorgeschlagen. Das System verwendet fehlerrobuste Videocodierung mit geringer Verzögerung und verzögerungssensitive Rückübertragungen, wobei mehrdimensionale Videoanpassung und gemeinsame Quellen- und Kanalressourcenzuteilung für bessere Kanalanpassung formuliert sind. Eine Full-Reference Videoqualitätsmetrik wird für die QoE-Schätzung in Gegenwart von räumlichen und zeitlichen Beeinträchtigungen entwickelt und ins System integriert. Experimentelle Ergebnisse zeigen, dass das QoE-orientierte System deutlich verbesserte Nutzerzufriedenheit für eine Vielzahl von Video-Inhalten und Kanalbedingungen liefert.