Search

Extreme climatic events such as drought and high temperatures are frequently occurring under the global climate change scenarios causing severe threats to crop yield. Several studies have used individual events, e.g., meteorological drought (dry event), high temperatures (hot event), or compound dry–hot events as predictors of agricultural drought severity. However, only a few studies considered the situation when at least one dry event or hot event appeared. Moreover, the ability to depict drought phenomena, in terms of duration and severity, with the same drought index for different time scales was often inconsistent. This study adopted the meta-Gaussian model (MG) to predict agricultural drought severity for the crops and vegetation growing seasons (from April to September) from the year 1948 to 2014 in China by using two predictors, namely, standardized dry or hot events index (SDHEI) and joint soil moisture deficit index (JSDI). The SDHEI was built based on monthly precipitation and daily temperature, and the JSDI was calculated by combining 1-, 3-, 6-, 9-, and 12-month standardized soil moisture index (SSI) based on monthly root zone soil moisture. The prediction performance of the MG was assessed by employing the Brier skill score (BSS), Nash–Sutcliffe efficiency coefficient (NSE), and root mean square error (RMSE). The severe agricultural drought in the months of April–September of 2010 in China was used as a case study. A visual comparison of this event indicated that the lead JSDI predictions for 1–3 months showed a resembled behavioral pattern with corresponding historical observations for most of the areas of China; however, it had some disagreements in the spatial extent and severity of the agricultural drought in some of the regions. The MG had a great performance with BSS > 0.4, NSE > 0.6, and RMSE

Direct finite element analysis of continuous spot welding was implemented with reverse geometric models of the actual stamping parts that comprise the back-side panel. This reverse panel showed apparent gaps after assembly and yielded to the electrodes and fixtures during welding, causing significant deformations. We acquired the total residual deformation in continuous simulation through a 3-D comparison between the original panel and the final deformed panel. This total deformation indicates that the deformations in the two processes offset with each other. We also performed a spot welding experiment and measured the actual residual deformation to validate the total deformation in simulation. They show good agreement in terms of deformation trend and magnitude. Finally, we designed seven comparative simulations to investigate the influence of the clamping pressure, electrode force, and spot welding sequence on the residual deformation. The results show that the sole increase of clamping pressure causes a deformation reduction in plate P4, while that of electrode force causes a deformation increase in P3. When the welding sequences of processes 1 and 2 are both reversed, the plate with the lowest deformation changes from P4 to P3. The final deformations in parts with multiple processes and multiple spot welds result from nonlinear interaction among many factors. The influence of a single factor change on the final deformation will be balanced by the remaining factors. This is why the fluctuations in total deformation are not significant between the standard simulation and each of the seven comparative simulations. The value of this article lies in the parameters optimization of spot welding based on reverse engineering and direct finite element analysis.

Translation Quality Estimation is critical to reducing post-editing efforts in machine translation and to cross-lingual corpus cleaning. As a research problem, quality estimation (QE) aims to directly estimate the quality of translation in a given pair of source and target sentences, and highlight the words that need corrections, without referencing to golden translations. In this paper, we propose Verdi, a novel framework for word-level and sentence-level post-editing effort estimation for bilingual corpora. Verdi adopts two word predictors to enable diverse features to be extracted from a pair of sentences for subsequent quality estimation, including a transformer-based neural machine translation (NMT) model and a pre-trained cross-lingual language model (XLM). We exploit the symmetric nature of bilingual corpora and apply model-level dual learning in the NMT predictor, which handles a primal task and a dual task simultaneously with weight sharing, leading to stronger context prediction ability than single-direction NMT models. By taking advantage of the dual learning scheme, we further design a novel feature to directly encode the translated target information without relying on the source context. Extensive experiments conducted on WMT20 QE tasks demonstrate that our method beats the winner of the competition and outperforms other baseline methods by a great margin. We further use the sentence-level scores provided by Verdi to clean a parallel corpus and observe benefits on both model performance and training efficiency., Accepted by The Web Conference 2021

Background:Hypertension is a highly prevalent disorder. A nomogram to estimate the risk of hypertension in Chinese individuals is not available.Methods:6201 subjects were enrolled in the study and randomly divided into training set and validation set at a ratio of 2:1. The LASSO regression technique was used to select the optimal predictive features, and multivariate logistic regression to construct the nomograms. The performance of the nomograms was assessed and validated by AUC, C-index, calibration curves, DCA, clinical impact curves, NRI, and IDI.Results:The nomogram140/90 was developed with the parameters of family history of hypertension, age, SBP, DBP, BMI, MCHC, MPV, TBIL, and TG. AUCs of nomogram140/90 were 0.750 in the training set and 0.772 in the validation set. C-index of nomogram140/90 were 0.750 in the training set and 0.772 in the validation set. The nomogram130/80 was developed with the parameters of family history of hypertension, age, SBP, DBP, RDWSD, and TBIL. AUCs of nomogram130/80 were 0.705 in the training set and 0.697 in the validation set. C-index of nomogram130/80 were 0.705 in the training set and 0.697 in the validation set. Both nomograms demonstrated favorable clinical consistency. NRI and IDI showed that the nomogram140/90 exhibited superior performance than the nomogram130/80. Therefore, the web-based calculator of nomogram140/90 was built online.Conclusions:We have constructed a nomogram that can be effectively used in the preliminary and in-depth risk prediction of hypertension in a Chinese population based on a 10-year retrospective cohort study.Funding:This study was supported by the Hebei Science and Technology Department Program (no. H2018206110).

Renal fibrosis is characterized by glomerulosclerosis and tubulointerstitial fibrosis in diabetic nephropathy (DN). We aimed to evaluate the effects of PP2 on renal fibrosis of DN. GSE33744 and GSE86300 were downloaded from the GEO database. Firstly, 839 DEGs were identified between nondiabetic and diabetic mice renal glomerular samples. COX-2 was selected to assess the effects of PP2 on renal glomerulosclerosis. In db/db mice, PP2 decreased the expression of COX-2, phosphorylated p65, and fibrotic proteins, accompanied with attenuated renal glomerulosclerosis. In cultured glomerular mesangial cells, high glucose- (HG-) induced p65 phosphorylation and COX-2 expression were attenuated by PP2 or NF-κB inhibitor PDTC. PP2, PDTC, or COX-2 inhibitor NS-398 ameliorated abnormal proliferation and expression of fibrotic proteins induced by HG. Secondly, 238 DEGs were identified between nondiabetic and diabetic mice renal cortex samples. UCP2 was selected to assess the effects of PP2 on renal tubulointerstitial fibrosis. In db/db mice, PP2 decreased the expression of PPARγ and UCP2, accompanied with attenuated renal tubulointerstitial fibrosis and EMT. In cultured proximal tubular cells, HG-induced PPARγ and UCP2 expression was inhibited by PP2 or PPARγ antagonist GW9662. PP2, GW9662, or UCP2 shRNA ameliorated HG-induced EMT. These results indicated that PP2 ameliorated renal fibrosis in diabetic mice.

The competitive performance of neural machine translation (NMT) critically relies on large amounts of training data. However, acquiring high-quality translation pairs requires expert knowledge and is costly. Therefore, how to best utilize a given dataset of samples with diverse quality and characteristics becomes an important yet understudied question in NMT. Curriculum learning methods have been introduced to NMT to optimize a model's performance by prescribing the data input order, based on heuristics such as the assessment of noise and difficulty levels. However, existing methods require training from scratch, while in practice most NMT models are pre-trained on big data already. Moreover, as heuristics, they do not generalize well. In this paper, we aim to learn a curriculum for improving a pre-trained NMT model by re-selecting influential data samples from the original training set and formulate this task as a reinforcement learning problem. Specifically, we propose a data selection framework based on Deterministic Actor-Critic, in which a critic network predicts the expected change of model performance due to a certain sample, while an actor network learns to select the best sample out of a random batch of samples presented to it. Experiments on several translation datasets show that our method can further improve the performance of NMT when original batch training reaches its ceiling, without using additional new training data, and significantly outperforms several strong baseline methods., Comment: Accepted as full paper by AAAI-2020 (oral presentation)

This paper presents our submissions to CCMT 2020 Quality Estimation (QE) sentence-level task for both Chinese-to-English (ZH-EN) and English-to-Chinese (EN-ZH). We propose new methods based on the predictor-estimator architecture. For the predictor, we propose XLM-predictor and Transformer-predictor. XLM-predictor novelly produces two kinds of contextual token representation, i.e., mask-XLM and non-mask-XLM. For the estimator, both RNN-estimator and Transformer-estimator are conducted and two novel strategies, i.e. top-K strategy and multi-head attention strategy, are proposed to enhance the sentence feature representation. We also propose new effective ensemble technique for sentence-level predictions.

Brachytherapy is an important radio-therapeutic modality for a variety of malignancies, including prostate cancer, cervix cancer, breast cancer, vagina cancer, endometrium cancer, head and neck cancer, and many more. This technique has been shown to be an effective and safe non-pharmaceutical treatment with fewer serious complications and better outcome than other treatments for breast cancer. Every year, hundreds of thousands of patients around the world benefit from brachytherapy, which reliably delivers a relatively higher radiation dose to the intended target. However, the follow-up time, patient eligibility criteria, treatment strategy, and radiation doses used in published studies are somewhat inconsistent, making it difficult to strictly compare and evaluate the performance of the treatment. More rigorous studies are required to confirm the safety of this technique and to make outcome data more comparable. In this review, we focus on recent advances in breast brachytherapy techniques and provide an overview of outcomes, cosmetic outcome, toxicity, complications, and limitations of brachytherapy for the treatment of breast cancer. We also summarize the clinical outcomes and toxicity results in patients receiving or not receiving brachytherapy.

Organizations are always confronted with the challenge of eciently nding out suitable candidates from massive re- sumes. Traditional human resource management based on the information management system usually adopts SQL queries or keywords search, which cannot capture the im- plicit information, while the manual work is always time- consuming. To ll this gap, this paper presents SmartHR, a resume query and management system based on seman- tic web. Beneting from knowledge base, it can understand users' intentions more intelligently and search for suitable candidates more accurately. In this paper, we propose two key technical diculties which SmartHR meets, including the complexity of knowledge base construction and the time- consuming semantic search, and then give appropriate solu- tions respectively. Four channels are adopted to construct knowledge base, which are well illustrated. Furthermore, a variety of performance optimizations are employed and the eectiveness is evaluated on real datasets of up to million- s of triples and the results show a great improvement. As a representative application in semantic web, our practice in SmartHR provides useful experience and conclusions for developers.

Small- and Medium-sized Enterprises benefits much from Service-oriented manufacturing, which utilizes the internet and service platform to arrange manufacturing resource and provides service according to the customers’ demands. On this platform, service selection is one of the key steps for customers to get the best services. This paper introduces a service selection approach in cloud manufacturing for Small- and Medium-size Enterprises. First, we build a service selection model, including service evaluation and service constraints. And then a service selection algorithm is presented based on the service selection model to identify the best service for a service buyer. Finally, a case study is given to illustrate how this approach works in Cloud Manufacturing Platform for Small- and Medium-sized Enterprises (CMfg-SME).

Rule execution is the core step of rule-based semantic web reasoning. However, most existing approaches are centralized, which cannot scale out to reason big semantic web datasets. In this paper, we described a kind of semantic web rule execution mechanism using MapReduce programming model, which not only can handle RDFS and OWL ter Horst semantic rules, but also can be used in SWRL reasoning. Theoretical analysis is present on the scalability of this rule execution mechanism. Result shows that it can scale well as Mapreduce framework.

A kind of analysis method in which SVM is used for power transformer DGA is proposed in this paper. This method uses the SVM algorithm to classify the composition of DGA in transformer and diagnoses the fault of the transformer. At the same time, it introduces the fuzzy membership function, and it can eliminate unable diagnosis area when the discrete decision function is used. Then by using a example to test this method, it shows the SVM play excellent performance in the fault diagnosis of power transformer.

With the market economic development of our country, the image design and brand consciousness comes into all aspects in our life. Nowadays, in the higher education, because it develops rapidly, the universities? competitions are intense. University is similar to an enterprise more and more, needs to to carry on the university image design and strengthen the brand construction. It is really important for us to find the development target accurately and take effective measures to portray the brand images.