Search

Machine learning is one of the most potential ways to realize the function prediction of the incremental large-scale G-protein-coupled receptors (GPCR). Prior research reveals that the key to determining the overall classification accuracy of GPCR is extracting valuable features and filtering out redundancy. To achieve a more efficient classification model, we put the feature synonym problem into consideration and create a new method based on functional word clustering and integration. Through evaluating the evolution correlation between features using the transition scores in mature molecular substitution matrices, candidate features are clustered into synonym groups. Each group of the clustered features is then integrated and represented by a unique key functional word. These retained key functional words are used to form a feature knowledge base. The original GPCR sequences are then transferred into feature vectors based on a feature re-extraction strategy according to the features in the knowledge base before the training and testing stage. We create multiple machine learning models based on Naïve Bayesian (NB), random forest (RF), support vector machine (SVM), and multi-layer perceptron (MLP) algorithms. The established model is applied to classify two public data sets containing 8354 and 12,731 GPCRs, respectively. These models achieve significant performance in almost all evaluation criteria in comparison with state-of-the art. This work demonstrated the potential of the novel feature extraction strategy and provided an effective theoretical design for the hierarchical classification of GPCRs.

The prestressed column, where unbonded posttensioned tendons are arranged along the axis of the RC column, showed self-centering behavior after lateral loading. To investigate the detailed hysteret...

The performance of deep neural networks is heavily dependent on its architecture and various neural architecture search strategies have been developed for automated network architecture design. Recently, evolutionary neural architecture search (EvoNAS) has received increasing attention due to the attractive global optimization capability of evolutionary algorithms. However, EvoNAS suffers from extremely high computational costs because a large number of performance evaluations are usually required in evolutionary optimization, and training deep neural networks is itself computationally very expensive. To address this issue, this article proposes a computationally efficient framework for the evolutionary search of convolutional networks based on a directed acyclic graph, in which parents are randomly sampled and trained on each mini-batch of training data. In addition, a node inheritance strategy is adopted so that the fitness of all offspring individuals can be evaluated without training them. Finally, we encode a channel attention mechanism in the search space to enhance the feature processing capability of the evolved neural networks. We evaluate the proposed algorithm on the widely used datasets, in comparison with 30 state-of-the-art peer algorithms. Our experimental results show that the proposed algorithm is not only computationally much more efficient but also highly competitive in learning performance.

As an important positioning source of indoor positioning technology, Wi-Fi signals have attracted the attention of researchers for a long time. Fingerprint positioning can solve the problems caused by non-line-of-sight propagation and multipath effects. To improve the accuracy of Wi-Fi indoor positioning, this study proposes an indoor positioning algorithm based on fine-grained channel state information (CSI) and convolutional neural network (CNN). CSI is a kind of observable measurement that better describes the nature of Wi-Fi signal propagation than received signal strength indication. This method uses the subcarrier amplitude and phase difference information extracted from CSI data to establish fingerprints. The clustering method is used to analyse the number of clusters of fingerprint data, and the fingerprint database is divided into two sub-databases according to the threshold. CNNs with the same network structure are used to train the two kinds of fingerprint sub-databases. In the positioning stage, the sub-database to which the data to be measured belongs is determined according to the calibration algorithm, and the corresponding CNN model is used to estimate the position. Experiments are performed in a typical indoor environment. Compared with existing fingerprint-based positioning methods, this method has higher positioning accuracy.

Millimetre wave radar is an emerging technology that can monitor vital signs without contact. This unique feature is very suitable for some particular situations, such as burn patient monitoring. C...

This paper investigates flight delay propagation in air transportation networks (ATNs) by considering both network structures and airport operation performance. An airport susceptible-infected-recovered (ASIR) model is established based on the mechanism of epidemic spreading, where the focus is on the impact of the infection rate in order to properly map and understand the probability of delay propagation. Different network configurations are abstracted under complex network theory, in which the ASIR model can be simulated upon. The simulation results show that the original airport traffic, airport connection and the level of airport turnaround services play important roles in influencing delay propagation in different airports. In addition, changes of network structure such as the emerging of secondary hubs can also influence the delay propagation.

Destination related applications are playing an important role in Internet of Vehicles(IoVs), which can provide people with convenience or business profit, such as traffic jam warning or parking guide. However, in reality, people hesitate to share their destination information to other people due to operation inconvenience, which requires service providers to predict vehicles' destinations in advance in order to deliver them destination related messages. Some papers have considered the delivery scheduling problem of destination related information. But, they neglect the destination prediction problem with the assumption that vehicle's destinations are known in advance. In this paper, we target the delivery scheduling problem of destination related information in the case of destinations unknown to others in IoVs. First, a realtime destination prediction framework with machine learning models is proposed, with which a vehicle's destination can be predicted while traveling. Then, we propose a delivery profit maximization algorithm for service providers to select a proper location to deliver destination related information to each vehicle. Simulations with real vehicle trajectories show that our scheduling algorithm performs well and can successfully select a proper location to disseminate destination related information.

Radar based microwave imaging system is proposed and used for breast cancer detection. Tumour location can be detected based on high dielectric differences between tumour and surrounding tissues. In clinical medicine, malignant tumour usually has more irregular morphology than benign tumour. However, previous work has ignored the discrimination between benign and malignant tumours, and this issue is solved in our work. In this paper, two major novel contributions are presented. Firstly, a complex natural responses (CNRs) method is proposed for discrimination between the benign and malignant tumours. The CNRs of the tumour show a unique feature that the CNRs only depend on the morphological appearance of the tumour. Simulation results show that the use of the CNRs can successfully discriminate between a 5 mm radius spherical benign tumour and a 5 mm average radius malignant tumour with a spiculated periphery. Secondly, microwave images of the benign and malignant tumours are created by a proposed mono-static imaging system which uses a single Vivaldi antenna to scan the breast. Both benign and malignant tumour positions can be correctly indicated from the images.

Likelihood probability based phylogenetic analysis approaches have contributed to impressive advances in minimizing the variance of estimating the evolutionary parameters. However, their actual applications are greatly limited due to the very time-consuming calculations of Conditional Likelihood Probabilities (CLPs). Accurately and quickly obtaining the likelihoods of massive tree samples can facilitate phylogenetic analysis process. Inspired by recent advance of machine learning techniques that greatly improve the performance of many related prediction tasks, this study proposes a Random Forest (RF) based learning and prediction approach, called NeoPLE. The approach initially learns the deep neighbor information between nodes from the topology representations of evolution trees, integrates likelihood information from these trees, and trains a non-linear prediction model. Instead of having to depend on the recursive calculations of the CLPs of tree nodes, NeoPLE transfers the process to a prediction by the trained model, thus the likelihood estimates become irrelevant with the calculations of CLPs. In terms of performance improvement, speedup factors ranging from 2.1 to 3.5X can be achieved on the analysis of realistic data sets. Moreover, NeoPLE is very suitable to handle the data sets having relatively large number of alignment sites, the factor of up to 27.5X can be achieved on the analysis of simulated data sets. In addition, NeoPLE is robust against a wide range of choices of evolutionary models and is ready to integrate in more phylogenetic inference tools. This study fills in the gaps of phylogenetic analysis using a machine learning approach in feature representation and likelihood prediction of evolution trees, which has not been reported in literatures.

Healthy lifestyle trends are getting more prominent around the world. There are numerous numbers of marathon running race events that have been held and inspired interest among peoples of different ages, genders and countries. Such diversified truths increase more difficulties to comprehending large numbers of marathon images, since such process is often done manually. Therefore, a new approach for racing bib number (RBN) localization and recognition for marathon running races using deep learning is proposed in this paper. Previously, all RBN application systems have been developed by using image processing techniques only, which limits the performance achieved. There are two phases in the proposed system that are phase 1: RBN detection and phase 2: RBN recognition. In phase 1, You Only Look Once version 3 (YOLOv3) which consists of a single convolutional network is used to predict the runner and RBN by multiple bounding boxes and class probabilities of those boxes.YOLOv3 is a new classifier network that outperforms other state-of-art networks. In phase 2, Convolutional Recurrent Neural Network (CRNN) is used to generate a label sequence for each input image and then select the label sequence that has the highest probability. CRNN can be straight trained from sequence labels such as words without any annotation of characters. Therefore, CRNN recognizes the contents of RBN detected. The experimental results based on mean average precision (mAP) and edit distance have been analyzed. The developed system is suitable for marathon or distance running race events and automates the localization and recognition of racers, thereby increasing efficiency in event control and monitoring as well as post-processing the event data

Graph Convolutional Networks (GCNs) are powerful representation learning methods for non-Euclidean data. Compared with the Euclidean data, labeling the non-Euclidean data is more expensive. Meanwhile, most existing GCNs only utilize few labeled data but ignore most of the unlabeled data. To address this issue, we design a novel end-to-end Iterative Feature Clustering Graph Convolutional Networks (IFC-GCN) that enhances the standard GCN with an Iterative Feature Clustering (IFC) module. The proposed IFC module constrains node features iteratively based on the predicted pseudo labels and feature clustering. Further, we design an EM-like framework for IFC-GCN training, which improves the network performance by rectifying the pseudo labels and the node features alternately. Theoretical analysis and experimental results show that our proposed IFC module can effectively modify the node features. Experimental results on public datasets demonstrate that IFC-GCN outperforms state-of-the-art methods on the semi-supervised node classification task.

In the study, our group choose a set of quality of red wine as data set. To get a more accurate result, we turn the quality into binary classification. And we try to build models to predict the quality of red wine based on machine learning algorithms, including Decision Tree, Boosting, Classification and regression tree and Random Forest. Among them, CART and Random Forest both get a high accuracy. A binary tree is built with CART and feature importance is analyzed. Meanwhile, we try to combine logistic algorithm with Random Forest and compare the accuracy of different models. In this way, it's found that there is a way to improve the accuracy of these models.

In recent years, deep convolutional neural networks (CNNs) have been widely used for hyperspectral image (HSI) classification. Besides, ensemble learning is a useful way to enhance the classification performance. Therefore, in this study, a new method titled Boosting-CNN is proposed for HSI classification, which fully explored the advantages of deep CNN and ensemble learning. Specifically, several deep CNNs are well-designed to classify HSI. The samples are misclassified in a CNN have more weighs in the following CNN through adaptive boosting. The final classification result is obtained by weighted voting of several CNNs. For HSI classification issue, the number of samples in different classes varies greatly, however, traditional classification methods cannot handle this issue well. In order to address imbalance training samples in HSI classification, soft class balanced loss is proposed to mitigate the influence of imbalance training samples. Experimental results on two popular hyperspectral datasets (i.e., Salinas and Pavia University) show that the proposed method obtain better classification accuracy compared to comparison methods.

Wildfires may cause great casualties and heavy wildfires are becoming more and more frequently all over the world in recent years. However, due to the environmental limitation, high manual-dependent operation is often impractical with other limits. In this paper, a transfer learning neural network based on long short term memory (LSTM) was used to detect wildfire based on Himawari-8. The real time dynamic threshold value detection for cloud mask based on the modified Otsu algorithm was used to fast and accurately remove cloud areas where wildfire detection is failed due to signal blocking. Then, the experiments were conducted with LSTM and other models. The experimental results showed that our method was positive for wildfire detection.

Jamming is a big threat to the survival of a radar system. Therefore, the recognition of radar jamming signal type is a part of radar countermeasure. Recently, convolutional neural networks (CNNs) have shown their effectiveness in radar signal processing, including jamming signal recognition. However, most of existing CNN methods do not regard radar jamming as a complex value signal. In this study, a complex-valued CNN (CV-CNN) is investigated to fully explore the inherent characteristics of a radar jamming signal, and we find that we can obtain better recognition accuracy using this method compared with a real-valued CNN (RV-CNN). CV-CNNs contain more parameters, which need more inference time. To reduce the parameter redundancy and speed up the recognition time, a fast CV-CNN (F-CV-CNN), which is based on pruning, is proposed for radar jamming signal fast recognition. The experimental results show that the CV-CNN and F-CV-CNN methods obtain good recognition performance in terms of accuracy and speed. The proposed methods open a new window for future research, which shows a huge potential of CV-CNN-based methods for radar signal processing.

This paper makes a detailed analysis of zero-voltage-switching (ZVS) realization in dual-active-bridge (DAB) converters based on extended-phase-shift (EPS) modulation, when the voltage ratio is not matched with the designed transformer turn ratio. Based on the ZVS principle that the charge stored in parasitic capacitance should be totally released, a specific lowest switching current is necessary at critical switching instants to achieve ZVS. A 400 V/400 V-5 kW laboratory prototype was fabricated to verify that the ZVS can be realized with the lowest switching current control, in which the switching current can be corrected as reference value at some instants. As a result, efficiency can be effectively increased in a specific power range with ZVS.

Chordomas are primary bone tumors that arise in the cranial base, mobile spine, and sacrococcygeal region, affecting patients of all ages. Currently, there are no approved agents for chordoma patients. Here, we evaluated the anti-tumor efficacy of small molecule inhibitors that target oncogenic pathways in chordoma, as single agents and in combination, to identify novel therapeutic approaches with the greatest translational potential. A panel of small molecule compounds was screened in vivo against patient-derived xenograft (PDX) models of chordoma, and potentially synergistic combinations were further evaluated using chordoma cell lines and xenograft models. Among the tested agents, inhibitors of EGFR (BIBX 1382, erlotinib, and afatinib), c-MET (crizotinib), and mTOR (AZD8055) significantly inhibited tumor growth in vivo but did not induce tumor regression. Co-inhibition of EGFR and c-MET using erlotinib and crizotinib synergistically reduced cell viability in chordoma cell lines but did not result in enhanced in vivo activity. Co-inhibition of EGFR and mTOR pathways using afatinib and AZD8055 synergistically reduced cell viability in chordoma cell lines. Importantly, this dual inhibition completely suppressed tumor growth in vivo, showing improved tumor control. Together, these data demonstrate that individual inhibitors of EGFR, c-MET, and mTOR pathways suppress chordoma growth both in vitro and in vivo. mTOR inhibition increased the efficacy of EGFR inhibition on chordoma growth in several preclinical models. The insights gained from our study potentially provide a novel combination therapeutic strategy for patients with chordoma. © 2021 The Authors. The Journal of Pathology published by John Wiley & Sons, Ltd. on behalf of The Pathological Society of Great Britain and Ireland.

With the economic development and people's growing demand for a better life, the proportion of home decoration market in the economic market is also gradually rising. However, under the impact of the downturn of the general environment and the new mode of Internet, many offline home decoration companies cannot bear the heavy burden and are struggling in the market operation. For a variety of marketing methods of businesses, consumers have immunity, little effect, so home decoration technology needs to produce a new model in the Internet environment. Though combining virtual reality technology with interior design, a new intelligent automation interior design system based on VR technology is introduced. This paper introduces the design and implementation of virtual system for interior design, and analyzes the key technologies of VR application. The main design idea of this system is to automatically design a perfect indoor layout with lighting effect, reflection effect and visual effect as the technical indicators according to different room types, including the functions of importing CAD drawings to dynamically create houses, automatically calculating scene layout, and accepting user interaction to change the design. In the specific implementation process, through the improved LOD technology, that is, according to the characteristics of VR scene, focus on rendering objects in the first person perspective, for all kinds of actors out of the perspective of high LOD optimization, PBR Technology (Physics Based Rendering Technology), combined pre-rendering and real-time Rendering can not only avoid the influence of the first person perspective effect caused by the general global rendering of the scene, but also avoid the high cost of computer performance caused by the rendering process.

Face verification has come into increasing focus in various applications including the European Entry/Exit System, which integrates face recognition mechanisms. At the same time, the rapid advancement of biometric authentication requires extensive performance tests in order to inhibit the discriminatory treatment of travellers due to their demographic background. However, the use of face images collected as part of border controls is restricted by the European General Data Protection Law to be processed for no other reason than its original purpose. Therefore, this paper investigates the suitability of synthetic face images generated with StyleGAN and StyleGAN2 to compensate for the urgent lack of publicly available large-scale test data. Specifically, two deep learning-based (SER-FIQ, FaceQnet v1) and one standard-based (ISO/IEC TR 29794-5) face image quality assessment algorithm is utilized to compare the applicability of synthetic face images compared to real face images extracted from the FRGC dataset. Finally, based on the analysis of impostor score distributions and utility score distributions, our experiments reveal negligible differences between StyleGAN vs. StyleGAN2, and further also minor discrepancies compared to real face images.

In order to overcome the shortcomings of the current higher education evaluation system, this paper starts with the current situation of higher education, and puts forward a new higher education evaluation system combining NET (national level, education level and comprehensive evaluation). The evaluation system refers to the 2010 higher education evaluation system of four countries, namely the United States, the United Kingdom, China and Japan. In addition, the evaluation system also constructs a system dynamics model to evaluate the education system of the four countries, and uses the linear regression method to calculate different variables, and then simulates and forecasts the data changes in the next ten years. The results show that the evaluation system can deeply study the shortcomings of the current higher education evaluation system, and can accurately predict the data changes of the higher education evaluation system in the next ten years, which undoubtedly lays a good foundation for improving the level of higher education evaluation and perfecting the higher education evaluation system in China.

Drip irrigation has recently been proposed as a water-saving method for rice cultivation. Growers have observed that rice yields are less under drip irrigation than under flood irrigation; however, the reason is unclear. The objective of this two-factor experiment was to compare the effects of irrigation method (flood irrigation and drip irrigation) and soil temperature (18, 24, and 30 °C) on rice physiology and yield. The results showed that at 30 °C soil temperature, drip irrigation reduced the photosynthetic activity of rice compared with flood irrigation. Drip irrigation also (i) increased the activities of superoxide dismutase by 5% and catalase by 27% and (ii) reduced endogenous hormone concentrations. At 18 °C soil temperature, drip irrigation significantly increased SOD (+ 7%) and CAT (+ 23%) and significantly reduced GA (− 25%) and IAA (− 16%) concentrations compared with flood irrigation. The ZR+Z concentrations were 13% less in drip irrigation than in flood irrigation at low soil temperature. The negative effects of low soil temperature (18 and 24 °C) were greater under drip irrigation than under flood irrigation, suggesting that low soil temperature exacerbated the stress that rice was under in the drip irrigation system. Drip irrigation significantly reduced the secondary spikelet number compared with flood irrigation. This led to a significant yield reduction. Overall, the combination of low soil temperature and drought stress may be the main reason for the marked decline in the yield of drip-irrigated rice in northern Xinjiang Province.

In this paper, a parallel model adaptive Kalman filtering algorithm is presented for multiple sensors estimation fusion when the measurement noise statistics are uncertain. As a typical adaptive filtering algorithm, the multiple model adaptive estimation tries to reduce the dependency of the filter on the noise parameters. It utilizes multiple models with different noise levels to estimate the state and combines the model-dependent estimates with model probability. However, with the increase in the number of active sensors, a large number of models are required to cover the entire range of the possible noise parameter values, which can become computationally infeasible. The main goal of this work is to incorporate the noise statistic estimator in the framework of the multiple model adaptive estimation, such that only two models are required for each sensor, which significantly reduce the complexity of the estimator. The advantage of the presented algorithm to deal with the model uncertainty is studied analytically. The high performance of the parallel model adaptive Kalman filtering for autonomous satellite navigation using inter-satellite line-of-sight measurements is illustrated in comparison with a robust Kalman filter, an intrinsically Bayesian robust Kalman filter, and the traditional multiple model adaptive estimation.

Purpose To answer 2 questions: What is the main structure that prevents the superior translation of the humeral head, the supraspinatus or the superior capsule (SC)? And what mechanism does the principal structure rely on to prevent the superior translation of the humeral head, the spacer effect or the tensional hammock effect? Methods Eight shoulder specimens were assessed using a custom biomechanical testing system. Glenohumeral superior translation and subacromial peak pressure were compared using 6 models: the intact joint model, supraspinatus dysfunction model, supraspinatus defect model, SC tear model, SC defect model, and irreparable rotator cuff tear (IRCT) model. Results Compared with the intact joint model, the supraspinatus defect model significantly increased the superior translation (by 2.6 mm; P .05). Conclusions The anatomic SC has a negligible role in preventing the superior translation of the humeral head. Clinical Relevance SC reconstruction is not a simple anatomic reconstruction, and its promising clinical outcome may be due to tensional fixation technique and choice of graft.

To achieve non-uniform imaging with a large field of view and high efficiency as well as to obtain an adjustable fovea with super-resolution, we proposed a curved retina-like camera array imaging system (CRCS), which is built by an eight-camera array distributed non-uniformly on a curved surface and a camera coaxial with Risley prisms located in the center. By the non-uniform imaging, the field of view of the developed prototype is 150 ∘ × 40 ∘ with a reduction of data redundancy by 87.62%. Besides, the experimental results show that CRCS can obtain clear and sharp images of farther targets of interest around the fovea field of view with a constant focal length.

The purpose of this study is to achieve an accurate classification of medical images of pneumonia using transfer learning. Pulmonary radiographs can provide physicians with an excellent aid in the diagnosis of pneumonia, and this paper collects images from patients with common bacterial pneumonia, COVID-19 disease, and normal incidents pulmonary x-rays are used as a dataset for transfer learning. The deep convolutional neural network framework proposed in recent years can be used to extract features from the dataset and classify them to obtain models with high accuracy. The transfer learning-based approach can not only reduce the dependence of the network model on the dataset but also significantly reduce the time and hardware costs, and the trained models usually produce significant results as well. A total of 4,500 x-ray images are collected in this paper, including 1,500 confirmed images of confirmed COVID-19 disease, 1,500 images of confirmed bacterial pneumonia, and 1,500 images of normal conditions. This paper, VGG19, ResNet-50, and GoogLeNet are used as the basis for transfer learning. The results show that using transfer learning can effectively extract advanced features of pneumonia and construct a pneumonia medical image recognizer, which is trained to recognize the models. The accuracy rates is 93.67%, 91.73%, and 90.80%, respectively, and the model shows excellent generalization ability. According to the results, this study can assist medical professionals in the diagnosis of pneumonia condition and effectively shorten the diagnosis time.

Short sleep duration potentially affects the risk of developing asthma and asthmatic exacerbation. Our aim is to determine the associations between sleep duration with asthmatic phenotypes, fraction of exhaled NO (FeNO), blood eosinophils, and lung function among current asthmatics. A total of 558 individuals from the National Health and Nutrition Examination Survey during 2007-2012 were included in this study. Self-reported sleep duration was divided into three groups: short (⩽6 h), healthy (7-8 h) and long (⩾9 h). By using a multivariate generalized additive model with binomial or Poisson regression, the abovementioned associations were evaluated after adjustment for potential confounders. Our study population included 284 short sleepers, 240 healthy sleepers and 34 long sleepers. In multivariate binomial regression analysis, we found that short sleep duration was associated with increased risk of asthma with central obesity (adjusted OR = 0.58, 95%CI: 0.33-0.99, P = 0.047) compared with healthy sleep duration. Multivariate Poisson regression analyses revealed a significant increase in FeNO (adjusted OR = 1.19, 95%CI: 1.14-1.26, P < 0.01) and reduction of blood eosinophil percentage (adjusted OR = 0.86, 95%CI: 0.77-0.96, P < 0.01) in the healthy sleep duration group compared with the short sleep duration group. The individuals with healthy and long sleep duration harbored better lung function than those with short sleep duration, with the exception of FEV1/FVC. Our study suggested that short sleepers potentially increase the risk of asthma with central obesity and lung function impairment than healthy sleepers among current asthmatics. We also observed that short sleep duration is associated with a significant reduction of FeNO and increase of blood eosinophil percentage.

To solve prostatic hyperplasia in the elderly, a method of cystostomy with plasma bipolar resection was proposed. From January 2019 to March 2020, 42 patients with BPH who needed surgical treatment in the urological department were selected. Cystostomy was performed in bipolar TURP. The cystostomy group and robot group were divided into two groups. The surgical safety, surgical efficiency, complications, and nursing time between the two groups were compared. The results showed that the experimental and control groups’ RUV values were significantly lower than those before surgery. In comparison, the Qmax value was considerably higher than that before surgery. The difference was statistically significant ( P < 0.05 ), suggesting that the cystostomy group in bipolar TURP had more substantial improvement of dysuria, better recovery of detrusor function, and better prognosis. It was proved that, for BPH below 80 g, cystostomy could reduce the operation time, bladder irrigation time, catheter indwelling time, and postoperative hospital stay, improve the operation efficiency, and have the same effect on patients’ symptoms improvement, more excellent psychological support, and higher quality of life score. It is proved that plasma bipolar resection combined with cystostomy can effectively improve annual BPH surgery.

Self-driving photodetectors have attracted much attention due to the advantages of the small size, low power consumption, maintenance-free. Although there are many researches on asymmetric self-driving photodetectors, most of them focus on the experimental control of materials and geometry, and lack of systematic theoretical research on the modulation mechanism of the device performance. Here, by coupling the electromagnetic response and the dynamics of semiconductor carriers via finite element method, we take perovskite as an example to systematically and detailedly explore the effects of the asymmetric Schottky barrier and doping concentration of the semiconductor on the response performance of the device from internal perspectives such as the distribution of the internal potential. It is found that a high doping concentration leads to the low short-circuit current and open-circuit voltage of the self-driving photodetector. Through the optoelectronic simulation and optimization, the MAPbI 3 -based self-driving photodetector with Au/Ag Schottky-junction shows excellent photoelectric performance at a wide spectral range (300 – 800 nm). It shows a high responsivity of 0.35 A/W and a specific detectivity of (5$\times$10)10 Jones under the 700 nm-wavelength illumination.

Background: Depression and anxiety contribute to an increased global burden of disease and affect women more often than men. Despite evidence from experimental data showing a shared biological mechanism involved in both mental health disorders and female hormone-dependent cancers, results from epidemiological investigations remain inconsistent.Methods: We aim to understand a putative causal relationship between psychological distress and female malignancy by conducting a two-sample Mendelian randomization (MR) analysis. We used summary statistics from the hitherto largest genome-wide association studies (GWAS) performed in depression (Ncase=246,363), anxiety (Ncase=44,465), breast (Ncase=122,977) and ovarian (Ncase=25,509) cancer. We constructed strong instruments using the 102 depression-associated SNPs, the 6 anxiety-associated SNPs, and applied several MR approaches.Results: We found that genetic predisposition to depression significantly increased the risk of both overall breast cancer (OR [95%CI] = 1.10 [1.03-1.18]) and its estrogen receptor (ER)- subtype (1.12 [1.01-1.24]), while a borderline significance was observed for ER+ subtype (1.08 [0.99-1.18]). These findings were corroborated by our genetic correlation analysis where a significantly shared genetic basis was observed for depression and breast cancer. On the contrary, we did not identify any causal association of anxiety with breast cancer. None of the mental health traits were associated with the onset of ovarian cancer or its serous subtype. Sensitivity analyses using different sets of instruments revealed consistent results.Conclusions: Our findings suggest that poor mental health condition such as major depression disorder is likely to be causally associated with the development of breast cancer, providing evidence supporting for the potential deleterious consequence of mental illness on cancer onset.

Background: Past history of gallstones is associated with increased risk of gallbladder cancer in observational studies. We conducted complementary observational and Mendelian randomization (MR) analyses to determine whether history of gallstones is causally related to development of gallbladder cancer in an Indian population. Methods: To investigate associations between history of gallstones and gallbladder cancer, we used questionnaire and imaging data from a gallbladder cancer case–control study conducted at Tata Memorial Hospital, Mumbai, Maharashtra, India (cases = 1,170; controls = 2,525). We then used 26 genetic variants identified in a genome-wide association study of 27,174 gallstone cases and 736,838 controls of European ancestry in an MR approach to assess causality. The association of these genetic variants with both gallstones and gallbladder cancer was examined in the gallbladder cancer case–control study. Various complementary MR approaches were used to evaluate the robustness of our results in the presence of pleiotropy and heterogeneity, and to consider the suitability of the selected SNPs as genetic instruments for gallstones in an Indian population. Results: We found a strong observational association between gallstones and gallbladder cancer using self-reported history of gallstones [OR = 4.5; 95% confidence interval (CI) = 3.5–5.8] and with objective measures of gallstone presence using imaging techniques (OR = 2.0; 95% CI = 1.5–2.7). We found consistent causal estimates across all MR techniques, with ORs for gallbladder cancer in the range of 1.3–1.6. Conclusions: Our findings indicate a causal relationship between history of gallstones and increased risk of gallbladder cancer, albeit of a smaller magnitude than those found in observational analysis. Impact: Our findings emphasize the importance of gallstone treatment for preventing gallbladder cancer in high-risk individuals.

Aiming at the problem of weak signal and a lot of noise in non-contact vital signs detection, a band-pass filter is designed to effectively extract the breathing signal and heart signal, and the heart signal is classified by LSTM network to identify the patients with atrial fibrillation (AFib) disease. Firstly, the vital signs obtained by millimeter wave radar is processed by range FFT, then perform phase extraction and phase unwrapping on the range-bin, then the breathing signal and heart signal are extracted by using the cascaded second-order filter. Finally, the filtered peak information of the heart signal is used to generate Gaussian pulse to simulate the ECG signal, and the LSTM network trained by the instantaneous frequency and spectral entropy features extracted from the 2017 PhysioNet/CinC challenge data is used to classify signals. The experimental results show that the detection algorithm can effectively detect the breath and heartbeat of human body, and recognize the AFib signal.

The dedicated memory of hardware accelerators can be insufficient to store all weights and/or intermediate states of large deep learning models. Although model parallelism is a viable approach to reduce the memory pressure issue, significant modification of the source code and considerations for algorithms are required. An alternative solution is to use out-of-core methods instead of, or in addition to, data parallelism. We propose a performance model based on the concurrency analysis of out-of-core training behavior, and derive a strategy that combines layer swapping and redundant recomputing. We achieve an average of 1.52x speedup in six different models over the state-of-the-art out-of-core methods. We also introduce the first method to solve the challenging problem of out-of-core multi-node training by carefully pipelining gradient exchanges and performing the parameter updates on the host. Our data parallel out-of-core solution can outperform complex hybrid model parallelism in training large models, e.g. Megatron-LM and Turning-NLG., ACM/IEEE Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis (SC'20)

Olfactory dysfunction is one of the early symptoms seen in Parkinson’s disease (PD). However, the mechanisms underlying olfactory pathology that impacts PD disease progression and post‐mortem appearance of alpha‐Synuclein (α‐Syn) inclusions in and beyond olfactory bulb in PD remain unclear. It has been suggested that environmental toxins inhaled through the nose can induce inflammation in the olfactory bulb (OB), where Lewy body (LB) is the first to be found, and then, spread to related brain regions. We hypothesize that OB inflammation triggers local α‐Syn pathology and promotes its spreading to cause PD. In this study, we evaluated this hypothesis by intranasal infusion of lipopolysaccharides (LPS) to induce OB inflammation in mice and examined cytokines expression and PD‐like pathology. We found intranasal LPS‐induced microglia activation, inflammatory cytokine expression and α‐Syn overexpression and aggregation in the OB via interleukin‐1β (IL‐1β)/IL‐1 receptor type I (IL‐1R1) dependent signaling. In addition, an aberrant form of α‐Syn, the phosphorylated serine 129 α‐Syn (pS129 α‐Syn), was found in the OB, substantia nigra (SN) and striatum 6 weeks after the LPS treatment. Moreover, 6 weeks after the LPS treatment, mice showed reduced SN tyrosine hydroxylase, decreased striatal dopaminergic metabolites and PD‐like behaviors. These changes were blunted in IL‐1R1 deficient mice. Further studies found the LPS treatment inhibited IL‐1R1‐dependent autophagy in the OB. These results suggest that IL‐1β/IL‐1R1 signaling in OB play a vital role in the induction and propagation of aberrant α‐Syn, which may ultimately trigger PD pathology.

Aims To study how to effectively prevent or reduce renal injury caused by contrast agents in diabetic patients. Main methods Sprague Dawley (SD) rats were bred with a high-fat diet for eight weeks, then intraperitoneally injected with Streptozotocin (STZ) to prepare the diabetes model. Rats were treated with Iodixanol to prepare a contrast-induced acute kidney injury (CIAKI) model. Moreover, 3-methyladenine (3-MA), an autophagy inhibitor, was administrated to diabetic rats with or without Rapamycin treatment. Serum creatinine (SCr) and blood urea nitrogen (BUN) were examined using Biochemical detector. Kidney injury molecule-1 (KIM-1), N-acetyl-β-D-amino glycosidase (NAG) in urine, inflammatory and oxidative stress factors in serum were determined by ELISA. The expression level of ROS was quantified by immunofluorescence (IF). The protein expressions of Bax, BCl-2, LC3, Beclin1, mTOR and p70S6K in renal tissue were detected by Western blot. Key findings Rapamycin was demonstrated to improve renal injury induced by Iodixanol diabetic rats, decrease the levels of SCr, BUN, KIM-1, NAG, improve renal functions, reduce inflammatory response and oxidative stress injury, down-regulate Bax, while up-regulate BCl-2 and inhibit apoptosis. Moreover, Rapamycin could inhibit the phosphorylation of mTOR/p70S6K pathway-associated proteins, activate autophagy and increase the levels of LC3 and Beclin1. After treatment with 3MA, an inhibitor of mTOR/p70S6K signaling pathway, the protective effects of Rapamycin on CIAKI were weakened. Significance Rapamycin can alleviate renal injury induced by Iodixanol diabetic rats, and its regulatory mechanisms may be related to the regulation of mTOR/p70S6K signaling pathway and the activating autophagy.

The primary objective of face morphing is to combine face images of different data subjects (e.g. a malicious actor and an accomplice) to generate a face image that can be equally verified for both contributing data subjects. In this paper, we propose a new framework for generating face morphs using a newer Generative Adversarial Network (GAN) - StyleGAN. In contrast to earlier works, we generate realistic morphs of both high-quality and high resolution of 1024$\times$1024 pixels. With the newly created morphing dataset of 2500 morphed face images, we pose a critical question in this work. \textit{(i) Can GAN generated morphs threaten Face Recognition Systems (FRS) equally as Landmark based morphs?} Seeking an answer, we benchmark the vulnerability of a Commercial-Off-The-Shelf FRS (COTS) and a deep learning-based FRS (ArcFace). This work also benchmarks the detection approaches for both GAN generated morphs against the landmark based morphs using established Morphing Attack Detection (MAD) schemes., Accepted in IWBF 2020

(1) Background: The novel coronavirus disease 2019 (COVID-19) is a global public health emergency that has caused worldwide concern. Vast resources have been allocated to control the pandemic and treat patients. However, little attention has been paid to the adverse impact on mental health or effective mitigation strategies to improve mental health. (2) Purpose: The aim of this study was to assess the adverse impact of the COVID-19 outbreak on Chinese college students&rsquo, mental health, understand the underlying mechanisms, and explore feasible mitigation strategies. (3) Methods: During the peak time of the COVID-19 outbreak in China, we conducted longitudinal surveys of sixty-six college students. Structured questionnaires collected information on demographics, physical activity, negative emotions, sleep quality, and aggressiveness level. A mixed-effect model was used to evaluate associations between variables, and the mediating effect of sleep quality was further explored. A generalized additive model was used to determine the dose-response relationships between the COVID-19 death count, physical activity, and negative emotions. (4) Results: The COVID-19 death count showed a direct negative impact on general sleep quality (&beta, = 1.37, 95% confidence interval [95% CI]: 0.55, 2.19) and reduced aggressiveness (&beta, = &minus, 6.57, 95% CI: &minus, 12.78, &minus, 0.36). In contrast, the COVID-19 death count imposed not a direct but an indirect impact on general negative emotions (indirect effect (IE) = 0.81, p = 0.012), stress (IE = 0.40, p &lt, 0.001), and anxiety (IE = 0.27, p = 0.004) with sleep quality as a mediator. Moreover, physical activity directly alleviated general negative emotions (&beta, 0.12, 95% CI: &minus, 0.22, &minus, 0.01), and the maximal mitigation effect occurred when weekly physical activity was about 2500 METs. (5) Conclusions: (a) The severity of the COVID-19 outbreak has an indirect effect on negative emotions by affecting sleep quality. (b) A possible mitigation strategy for improving mental health includes taking suitable amounts of daily physical activity and sleeping well. (c) The COVID-19 outbreak has reduced people&rsquo, s aggressiveness, probably by making people realize the fragility and preciousness of life.

Wireless sensor network (WSN) consists of base stations and sensors nodes to monitor physical and environmental conditions. Power consumption is a challenge in WSN due to activities of nodes. High power consumption is required for the main transceiver in WSN to receive communication requests all the time. Hence, a low power wake-up receiver is needed to minimize the power consumption of WSN. In this work, a low power wake-up receiver using ultrasound data communication is designed. Wake-up receiver is used to detect wake-up signal to activate a device in WSN. Functional block modelling of the wake-up receiver is developed in Silterra CMOS 130nm process technology. The performance of the wake-up receiver has been analyzed and achieving low power consumption which is 22.45μW. A prototype to demonstrate a wireless sensor node with wake-up receiver has been developed incorporating both ultrasonic and RF for internal and external communication respectively. We achieve 99.97% of power saving for 10s operation in the experimental setup for the WSN with and without wake-up receiver. Wake-up receiver used in WSN save power and prolong the lifetime of batteries and thus extending the operational lifetime of WSN.

Federated learning is a recently proposed distributed machine learning paradigm for privacy preservation, which has found a wide range of applications where data privacy is of primary concern. Meanwhile, neural architecture search has become very popular in deep learning for automatically tuning the architecture and hyperparameters of deep neural networks. While both federated learning and neural architecture search are faced with many open challenges, searching for optimized neural architectures in the federated learning framework is particularly demanding. This survey paper starts with a brief introduction to federated learning, including both horizontal, vertical, and hybrid federated learning. Then neural architecture search approaches based on reinforcement learning, evolutionary algorithms and gradient-based are presented. This is followed by a description of federated neural architecture search that has recently been proposed, which is categorized into online and offline implementations, and single- and multi-objective search approaches. Finally, remaining open research questions are outlined and promising research topics are suggested.

The nonlinear relationship between influential factors and house price and inadequate number of sample size could be the cause of the poor performance of the traditional models. Meanwhile, the daily data of the real estate market is very huge and it is increasing rapidly. The traditional house price prediction approaches are lack of capacity for massive data analysis, causing low utilization of data. To address these concerns, a house price prediction model based on deep learning is proposed in this paper, implemented on the TensorFlow framework. Adam optimizer is used to train the model, where Relu function is adopted to be the activation function. Then house price trend is predicted based on the ARIMA model. By using Scrapy, housing data are obtained from Internet to be the experimental dataset. Comparative experiments were conducted between the proposed approach and SVR method. The experimental results show that individual house price predicted by the proposed approach is better than that of SVR method. And the predicted house price trend is mainly agreement with the real situation.

Gated recurrent unit (GRU) networks perform well in sequence learning tasks and overcome the problems of vanishing and explosion of gradients in traditional recurrent neural networks (RNNs) when learning long-term dependencies. Although they apply essentially to financial time series predictions, they are seldom used in the field. To fill this void, we propose GRU networks and its improved version for predicting trading signals for stock indexes of the Hang Seng Indexes (HSI), the Deutscher Aktienindex (DAX) and the S&P 500 Index from 1991 to 2017, and compare the GRU-based models with the traditional deep net and the benchmark classifier support vector machine (SVM). Experimental results show that the two GRU models proposed in this paper both obtain higher prediction accuracy on these data sets, and the improved version can effectively improve the learning ability of the model.

A lead-iodide single crystal semiconductor with multi-orientation photoconductive properties and broad spectrum absorption has been designed and constructed by a crystal engineering technique. A strong charge transfer band was formed via organic–inorganic intermolecular interactions, darkening the color of the crystal. Electron catcher viologen cations surrounding the inorganic [Pb2I6]2−n chains extended the conductive orientation from along the a-axis in its perpendicular direction, presenting an interesting multi-channel conductive single crystalline system.

Sprayed dust suppression techniques are widely used to reduce the dust concentration in mines, and evaporative cooling by water spray is increasingly used as an approach to enhance thermal comfort in built environments. In mines, the roadway under excavation, which is known as the development heading, has not yet been connected to through ventilation. Auxiliary ventilation is the most important but also the most difficult to achieve in this area. Operators face the dual threat of dust and high temperatures in the area, in particular at the end of the heading, that is, at the heading face. To improve the working environment in this area, the present study improved the load-haul-dump vehicle and installed an external spray system on the bucket of the vehicle. To verify the cooling and dust suppression effect of the system, a computational fluid dynamics simulation was adopted to simulate the velocity and temperature fields of the airflow and the dynamic distributions of the water droplets using the Euler–Lagrange approach. The study shows that the water droplets can completely cover the dust-producing area, and the water temperature of the spray system can significantly affect the temperature distribution on the airflow in the development heading.

Objective We aimed to identify the factors that can predict the risk of occipital lobe damage preoperatively when resecting tumors located at the tentorial or pineal regions with the occipital-transtentorial approach (Poppen approach). Methods In 27 consecutive patients who underwent tumor resection with the Poppen approach for tentorial or pineal region meningiomas, the following morphologic parameters were assessed on a preoperative magnetic resonance imaging: (1) tentorial angle, (2) tentorial length, and (3) the shortest distance from the confluence of the sinus to the tumor. These parameters, together with tumor size, texture, and resection extent, were correlated with occipital lobe damage by using the one-way analysis of variance, χ 2 , or Fisher's exact tests. Results The mean value was 55.3° ± 5.6° (range, 45°–66°) for the tentorial angle, which was significantly associated with the occipital lobe damage grades ( P = 0.008), but this was not the case for the tentorial length ( P = 0.802) and the shortest distance from the confluence of the sinus to the tumor ( P = 0.695). Interestingly, age was also strongly associated with occipital lobe damage risk ( P = 0.020). The patients in the subgroup with no occipital damage (grade 4) were the youngest (aged 47.3 years), compared with other grades, with age of 58.0 years for grade 1, 54.3 years for grade 2, and 58.6 years for grade 3. These 2 parameters were also significant after multivariate analysis. No correlation was observed between either tumor nature or the extent of resection and damage grades. Conclusions The risk of occipital lobe damage increases in the presence of a steep tentorial angle during the Poppen approach for tentorial or pineal area tumors. Awareness of such anatomic features preoperatively is important for minimizing operative complications.

Motivation Genome-wide association studies are now shifting focus from analysis of common to rare variants. As power for association testing for individual rare variants may often be low, various aggregate level association tests have been proposed to detect genetic loci. Typically, power calculations for such tests require specification of large number of parameters, including effect sizes and allele frequencies of individual variants, making them difficult to use in practice. We propose to approximate power to a varying degree of accuracy using a smaller number of key parameters, including the total genetic variance explained by multiple variants within a locus. Results We perform extensive simulation studies to assess the accuracy of the proposed approximations in realistic settings. Using these simplified power calculations, we develop an analytic framework to obtain bounds on genetic architecture of an underlying trait given results from genome-wide association studies with rare variants. Finally, we provide insights into the required quality of annotation/functional information for identification of likely causal variants to make meaningful improvement in power. Availability and implementation A shiny application that allows a variety of Power Analysis of GEnetic AssociatioN Tests (PAGEANT), in R is made publicly available at https://andrewhaoyu.shinyapps.io/PAGEANT/. Supplementary information Supplementary data are available at Bioinformatics online.

In recent years, China has experienced unprecedented growth in air travel demand. However, existing hub airports have not been able to satisfy the increasing traffic, resulting in serious congestion and/or flight delays. As a result, regional hub airports located near hub airports have been increasingly utilized in recent years. This paper examines the factors that influence the emergence of regional hub airports, and determines the potential of regional hubs by means of a strict capacitated single allocation of p-hub median locating problem model. The empirical investigation focuses on the routes between airports in Jiangsu Province and other major airports in China. One result is for instance that Nanjing Lukou International Airport is identified to be the optimal regional hub airport for Pudong International Airport, China’s busiest hub airport. Results also shows that utilizing a regional hub airport would enhance the network performance in terms of reduced total network cost and congestion at the hub...

As the scale of software and the complexity of programs continue to grow, it is hard to meet development of modern computer technology only by manually extracting program features. In recent years, deep learning has achieved rapid development in different fields. Program representation learning based on deep learning has been widely used in many works, such as software vulnerability analysis, program analysis and malware detection. And it has gradually become a hot research direction in information security. After the deep analysis of existing research work on automatic program security detection, we catalog deep program representation learning for program security based on data representation and provide a comprehensive overview of deep program representation learning for program security under different application scenes. Then, we propose a deep program representation learning framework for program security. Finally, we conduct comparative analysis and summarize the challenges in deep program representation learning for program security.

Vascular endothelial growth factor (VEGF) and its receptor (VEGFR) play a fundamental role in tumor growth and angiogenesis. Signalling through VEGF/VEGFR not only induces cancer angiogenesis but also affects immune cells, making VEGF/VEGFR a pair of well-documented cancer therapeutic targets. Recently more and more studies have focused on combination of anti-VEGF/VEGFR agents and immune checkpoint inhibitors in preclinical and clinical settings. BD0801 is a humanized rabbit anti-VEGF monoclonal antibody currently in clinical development stage. In this study, we have revealed that BD0801 presents more potent activity than Bevacizumab in blockade of VEGF/VEGFR2 binding (IC50=275 ng/ml for BD0801; IC50=1451 ng/ml for Bevacizumab) and VEGFR2 activation in HUVEC assays. BD0801 also showed enhanced inhibitory effects on the proliferation (IC50=87 ng/ml for BD0801; IC50=476 ng/ml for Bevacizumab) and migration of the HUVECs, compared to Bevacizumab. We found that BD0801 exhibits dose-dependent tumor growth inhibitory activities in both lung cancer PC9 xenograft and lung cancer 3LL murine syngeneic tumor models. In addition, we obtained PK parameters of BD0801 using lung cancer PC9 tumor bearing mice and successfully made correlations between the in vitro function, PK profile and in vivo efficacy of BD0801. Notably, combination of BD0801 with either anti-PD-1, or anti-PD-L1 antibodies showed synergistic anti-tumor efficacy in both lung and colorectal cancer mouse models. Mechanistic studies were conducted to reveal possible mechanism of actions (MOA) underlying the modification in tumor microenvironment. Our data demonstrated that the MOA of the anti-tumor synergy may involve enhanced T-cell mediated immunity, including increased tumor infiltration of CD8+ and CD4+ T cells and reduced double positive CD8+PD-1+ T cells, as well as improved vasculature normalization. Taken together, these data provide a foundation for combining BD0801 with immunotherapy for cancer treatment and support further clinical development plans for BD0801 toward this direction. Citation Format: Liting Xue, Haoyu Zhang, Qian Wang, Feng Li, Xinxin Li, Xiaohong Yu, Zhihong Lu, Yue Huang, Wenqing Yang. Combination of an anti-angiogenic antibody with PD1/PDL1 blockade agents produces synergistic anti-tumor efficacy in preclinical models [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 2651.