Search

Uncertainty in the food supply continues to increase due to COVID-19 and geopolitical conflicts, and reducing farmland abandonment is an important measure for ensuring food security. In recent years, the Chinese government has promoted e-commerce demonstration projects in rural areas, so we ask does the spread of e-commerce reduce farmland abandonment? This study first constructs a theoretical framework for the effect of e-commerce on farmland abandonment and then quantifies that effect through spatial analysis and econometric models based on 12,732 households in the China Rural Household Panel Survey (CRHPS). The study found that there was a significant spatial aggregation between the popularization of e-commerce and the distribution of abandoned farmland, with a significant negative correlation between these two factors. The popularization of e-commerce has reduced farmland abandonment. The probability and size of farmland abandonment by farmers dropped by 4.1 % and 17.5 % when e-commerce was promoted in a village; the probability and size of farmland abandonment by farmers dropped by 7.8 % and 38.7 % when the rate of the popularization for e-commerce in villages increased by 1 %, respectively. E-commerce can reduce farmland abandonment in rural areas by absorbing returning labour, stimulating the land transfer market and revitalizing rural industry. With the improvement of physical infrastructure such as rural roads, the government should pay attention to the construction of digital villages including e-commerce infrastructure and continue to innovate new ways small farmers to participate in e-commerce, thus helping transform rural spaces and reusing idle land resources. • There is a significant negative relationship between e-commerce and farmland abandonment. • The popularization of e-commerce has reduced farmland abandonment. • E-commerce can reconfigure factors of production by absorbing labour, stimulating land transfer and revitalizing industry. • Digital village construction deserves to be given attention to revitalize the unused land resources. [ABSTRACT FROM AUTHOR]

Biological signals, including electroencephalography (EEG) and electrooculography (EOG), are often used to develop switches, which represent a class of typical asynchronous human–computer interfaces (HCIs) in which control and idle states need to be distinguished based on a criterion. Determining a satisfactory criterion for rapid and accurate discrimination between control and idle states remains a challenging issue, as EEG signals are highly noisy and nonstationary, and EOG signals are highly affected by unintended/spontaneous eye movements. Therefore, most existing EEG- or EOG-based switches are characterized by disadvantages of long response times (RTs) or high false positive rates (FPRs). The primary contribution of this work is the development of a novel EOG-based switch design, in which a visual trigger mechanism is introduced to guide the users’ blinks and to assist in detecting blinks. Specifically, the graphical user interface (GUI) includes a switch button that flashes once per 1.2 s. The user is instructed to blink synchronously with the flashes of the switch button to issue an on/off command while a single-channel EOG signal is collected. A waveform detection algorithm is applied to the ongoing EOG signal, which discriminates the intended and unintended blinks mainly based on the synchrony between the blink and the switch buttons flash. Once an intended blink, i.e., a blink corresponding to a button’s flash, is detected, the system issues an on/off command. As one application, the proposed EOG-based switch is used to produce start/stop commands for a wheelchair. Several online experiments were conducted with ten healthy subjects. An average accuracy of 99.5%, an RT of 1.3 s for issuing a switch command in the control state, and an average FPR of 0.10/min in the idle state were achieved. The experimental results therefore demonstrate the effectiveness of the single-channel EOG-based switch. [ABSTRACT FROM AUTHOR]

Multivariate pattern analysis (MVPA) approach applied to neuroimaging data, such as functional magnetic resonance imaging (fMRI) data, has received a great deal of attention because of its sensitivity to distinguishing patterns of neural activities associated with different stimuli or cognitive states. Generally, when using MVPA approach to decode the mental states or stimuli, a set of discriminative variables (e.g., voxels) is first selected. However, in most of existing MVPA methods, the selected variables do not contain all informative variables, since these selected variables are sufficient for decoding. In this paper, we propose a multivariate pattern analysis method based on sparse representation for decoding the brain states and localizing category-specific brain activation areas corresponding to two experimental conditions/tasks at the same time. Unlike traditional MVPA approaches, this method is designed to find informative variables as many as possible. We applied the proposed method to two judgement experiments: a gender discrimination and an emotion discrimination task, data analysis results demonstrate its effectiveness and potential applications. [ABSTRACT FROM AUTHOR]

Low molecular weight, uncharged, halogenated disinfection byproducts (DBPs) are poorly removed by the reverse osmosis and advanced oxidation process treatment units often applied for further treatment of municipal wastewater for potable reuse. Granular activated carbon (GAC) treatment effectively sorbed 22 halogenated DBPs. Conversion of the GAC to a cathode within an electrolysis cell resulted in significant degradation of the 22 halogenated DBPs by reductive electrolysis at −1 V vs. Standard Hydrogen Electrode (SHE). The lowest removal efficiency over 6 h electrolysis was for trichloromethane (chloroform; 47%) but removal efficiencies were >90% for 13 of the 22 DBPs. In all cases, DBP degradation was higher than in electrolysis-free controls, and degradation was verified by the production of halides as reduction products. Activated carbons and charcoal were more effective than graphite for electrolysis, with graphite featuring poor sorption for the DBPs. A subset of halogenated DBPs (e.g., haloacetonitriles, chloropicrin) were degraded upon sorption to the GAC, even without electrolysis. Using chloropicrin as a model, experiments indicated that this loss was attributable to the partial reduction of sorbed chloropicrin from reducing equivalents in the GAC. Reducing equivalents depleted by these reactions could be restored when the GAC was treated by reductive electrolysis. GAC treatment of an advanced treatment train effluent for potable reuse effectively reduced the concentrations of chloroform, bromodichloromethane and dichloroacetonitrile measured in the column influent to below the method detection limits. Treatment of the GAC by reductive electrolysis at −1 V vs. SHE over 12 h resulted in significant degradation of the chloroform (63%), bromodichloromethane (96%) and dichloroacetonitrile (99%) accumulated on the GAC. The results suggest that DBPs in advanced treatment train effluents could be captured and degraded continuously by reductive electrolysis using a GAC-based cathode. [ABSTRACT FROM AUTHOR]

Abstract: The effects of ultrasonication pre-treatment of graphene oxide (GO) on the mechanical properties of GO/poly(vinyl alcohol) (PVA) composites were studied. Results show that the mechanical properties are sensitive to ultrasonication time (or energy input) and there is an optimum ultrasonication time (OUT) that leads to the largest improvement in mechanical property. If the ultrasonication time is less than OUT, the GO sheets are not fully exfoliated, and only a partial reinforcement effect is achieved. Ultrasonication times longer than OUT cause a reduction of GO sheet size and impair the efficiency of mechanical improvement. The optimized ultrasonication energy input of 15Wh/L may serve as a general guideline for preparation of GO composites. [Copyright &y& Elsevier]

Abstract: The growing demand for compact point-of-care medical devices and portable instruments for on-site environmental sampling has stimulated intense research on flexible sensors that can be miniaturized and function under considerable physical deformation. We report a new type of flexible electrochemical biosensors based on free-standing graphene paper carrying binary nanocomposites of PtAu alloy and MnO2. The coral-like PtAu–MnO2 nanocomposites are grown on the substrate through one-step template-free electrodeposition, leading to an intimate contact between the PtAu alloy and MnO2 matrix. The flexible electrode exhibits a unique set of structural and electrochemical properties such as better uniformity, larger active surface areas, and faster electron transfer in comparison with the control electrode prepared by tandem growth of MnO2 network and PtAu alloy in two steps. In nonenzymatic amperometric glucose detection, the PtAu–MnO2 binary nanostructure-decorated graphene paper has shown greatly enhanced sensing performance such as wide liner range (0.1mM to 30.0mM), high sensitivity (58.54μAcm−2 mM−1), low detection limit (0.02mM, S /N=3), satisfactory selectivity, excellent reproducibility and stability, and tolerability to mechanical stress. The strategy of co-growth of metal and metal oxides on freestanding carbon substrates opens new possibility to develop high-performance flexible electrochemical sensors. [Copyright &y& Elsevier]

Abstract: The immobilization of carboxylesterase from Spodoptera Litura in mesoporus molecular sieves was studied in this paper. Two different types of mesoporous sieves (MCM-41, SBA-15) were initially used as enzyme support to compare the efficiency of immobilization. The outcome showed the maximum enzyme loadings, twice as many in MCM-41, are 0.0267 L/g in SBA-15 support. The most efficient pH value for immobilization was at 6.5 for both types of the support and the optimal immobilizing time was 2h for SBA-15 and 4h for MCM-41. Hence, the SBA-15 mesoporous support was determined as the enzyme immobilization matrix to continue the investigation because of its superior properties for enzyme immobilization. Then a series of characters of the free and immobilized Carboxylesterase were compared. As a result, the immobilized enzyme was substantiated to be obviously more stable than the free enzyme under the variant conditions. The experiments of pesticides degradation show that, relative to degradation in natural environment, immobilization enzyme had a higher ability to reduce the organic compounds with ester bonds. [Copyright &y& Elsevier]

Abstract: A surface-draw method to fabricate recyclable carbon nanotube/polyvinyl butyral (CNT/PVB) composite fibers is reported. This method is effective for both single-walled carbon nanotube (SWCNT) and multi-walled carbon nanotube. The CNT mass content of CNT/PVB composite fibers can vary from 0 to 80wt.%, which is higher than most CNT/polymer composites reported to date. The diameter of the composite fibers can be controlled in the range of 10–100μm, with essentially unlimited draw length. The composite fibers with 7.4wt.% SWCNTs showed optimal tensile properties. Compared with pure PVB fibers, the tensile strength, failure strain, and elastic modulus of the composite fiber have improved about 127%, 27%, and 73%, respectively. In addition, SWCNT/PVB composites with 66.7wt.% SWCNTs have the highest conductivity of 42.9Sm−1. More importantly, the major benefit is the “greenness” of the method, which involves environment friendly ethanol–water solvent with no functionalization of the nanotube required, and only simple apparatus are needed. The CNT/PVB composite fibers obtained can be dissolved in ethanol solution and reformed with the surface draw method without any additional treatment; and the material properties after recycle is comparable to those fabricated in the first round. [Copyright &y& Elsevier]

Abstract: This paper discusses blind source extraction in various ill-conditioned cases based on a simple extraction network model. Extractability is first analyzed for the following ill-conditioned cases: the mixing matrix is square but singular, the number of sensors is smaller than that of sources, the number of sensors is larger than that of sources but the column rank of mixing matrix is deficient, and the number of sources is unknown and the column rank of mixing matrix is deficient. A necessary and sufficient condition for extractability is obtained. A cost function and an unsupervised learning algorithm for the extraction network model are developed. Simulation results are also presented to show the validity of the theoretical results and the performance and characteristics of the learning algorithm. [Copyright &y& Elsevier]

This paper discusses the bifurcation on the oscillation of the following second order ordinary differential equation:x″(t)+tλp(t)f(x)=0.Several conditions for the functions p(t), f(x) are obtained, under which the equation is nonoscillatory as λ<−1, and oscillatory as λ>−1, that is, λ=−1 is a bifurcation value on oscillation of the equation. [Copyright &y& Elsevier]

Cyclin-dependent kinase 9 (CDK9) is an increasingly important potential cancer treatment target. Nowadays, developing selective CDK9 inhibitors has been extremely challenging as its ATP-binding sites are similar with other CDKs. Here, we report that the CDK9 inhibitor BAY-1143572 is converted into a series of proteolysis targeting chimeras (PROTACs) which leads to several compounds inducing the degradation of CDK9 in acute myeloid leukemia cells at a low nanomolar concentration. In addition, the most potent PROTAC molecule B03 could inhibit cell growth more effectively than warhead alone, with little inhibition of other kinases. This enhanced antiproliferative activity is mediated by a slight increase in kinase inhibitory activity and an increase in the level of apoptosis induction. Moreover, B03 could induce the degradation of CDK9 in vivo. Our work provides evidence that B03 represents a lead for further development and that CDK9 degradation is a potential valuable therapeutic strategy in acute myeloid leukemia. Image 1 • A series of CDK9 PROTACs have been obtained through the conversion of a selective inhibitor BAY-1143572. • The most promising compound B03 has high kinase selectivity profiles. • Compound B03 remarkably inhibit the growth of acute myeloid leukemia cells. • B03 could degrade CDK9 effectively in vitro and in vivo. [ABSTRACT FROM AUTHOR]

A triple modular redundancy (TMR) based design technique for double cell upsets (DCUs) mitigation is investigated in this paper. This technique adds three extra self-voter circuits into a traditional TMR structure to enable the enhanced error correction capability. Fault-injection simulations show that the soft error rate (SER) of the proposed technique is lower than 3% of that of TMR. The implementation of this proposed technique is compatible with the automatic digital design flow, and its applicability and performance are evaluated on an FIFO circuit. [ABSTRACT FROM AUTHOR]

Abstract Accurately and effectively mapping and evaluating cultivated land fallow has already become an important issue that has received much attention in China. However, systematically analysing regional cultivated land fallow remains inadequate because current studies have mainly focused on quantifying cultivated land fallow using statistical data based on administrative units or a single aspect of cultivated land fallow using high or medium spatial resolution images at the local or regional scales. Against the existing shortcomings, this study first developed an integrated index of cultivated land fallow (ILF) for mapping and evaluating cultivated land fallow in Southwest China using multisource spatial data. The performance of the ILF was validated by comparing its results with Google Earth images and ecological carrying capacity of cultivated land (TEC). And the spatial distribution of cultivated land fallow in Southwest China was evaluated at the regional, provincial and metropolitan scales. The results revealed that the ILF provided a reliable evaluation of cultivated land fallow in Southwest China. Compared to the Google earth images, the pixel with the high ILF value was the cultivated land that was found to prioritize fallow. There was also a significant correlation between ILF and TEC at the prefectural level in Sichuan, with an R 2 value >0.65. In Southwest China, the cultivated land related to highly appropriate fallow (HAF) accounted for 5.73% of the total cultivated land in 2010. The cultivated land related to inappropriate fallow (IF) accounted for 53.26% and 37.36% in Sichuan and Chongqing but only comprised 22.90% and 19.72% in Yunnan and Guizhou, respectively. Special attention needs to be paid to Guiyang and Kunming, where the HAF made up 25.38% and 17.48% of their total cultivated land, respectively. Human activities have been found to already become the most important impact factors for cultivated land fallow in Southwest China. This study is especially valuable for providing a scientific basis for policy-making on viable cultivated land fallow policy in Southwest China. Graphical abstract Unlabelled Image Highlights • A simple and direct way was proposed to map cultivated land fallow in Southwest China. • The spatial distribution of cultivated land fallow in Southwest China was evaluated from multiscale perspectives. • Human activities are among the major impact factors of cultivated land fallow in Southwest China. [ABSTRACT FROM AUTHOR]

In recent years, various applications of brain computer interfaces (BCIs) have been studied. In this paper, we present a hybrid BCI combining P300 and motor imagery to operate an explorer. Our system is mainly composed of a BCI mouse, a BCI speller and an explorer. Through this system, the user can access his computer and manipulate (open, close, copy, paste, and delete) files such as documents, pictures, music, movies and so on. The system has been tested with five subjects, and the experimental results show that the explorer can be successfully operated according to subjects’ intentions. [ABSTRACT FROM AUTHOR]

Abstract: We present a new linear discriminant analysis method based on information theory, where the mutual information between linearly transformed input data and the class labels is maximized. First, we introduce a kernel-based estimate of mutual information with a variable kernel size. Furthermore, we devise a learning algorithm that maximizes the mutual information w.r.t. the linear transformation. Two experiments are conducted: the first one uses a toy problem to visualize and compare the transformation vectors in the original input space; the second one evaluates the performance of the method for classification by employing cross-validation tests on four datasets from the UCI repository. Various classifiers are investigated. Our results show that this method can significantly boost class separability over conventional methods, especially for nonlinear classification. [Copyright &y& Elsevier]

[Display omitted] Histone acetylation is one of the most essential parts of epigenetic modification, mediating a variety of complex biological functions. In these procedure, p300/CBP could catalyze the acetylation of lysine 27 on histone 3 (H3K27ac), and had been reported to mediate tumorigenesis and development in a variety of tumors by enhancing chromatin transcription activity. Ovarian cancer, as an extremely malignant tumor, has also been observed to undergo abnormal acetylation of histones. However, whether the treatment of ovarian cancer could be achieved by inhibiting the acetylation activity of p300/CBP on H3K27 has not been well investigated. In this article, we modified the structure of p300/CBP HAT domain inhibitor A-485 and obtained a highly active small molecule known as 13f , which has an IC 50 value of 0.49 nM for inhibiting the in vitro enzyme activity of p300, as well as the anti-proliferation IC 50 value on ovarian cancer cell line OVCAR-3 was 153 nM. In addition, 13f had strong acetylase family selectivity, good metabolic stability and promising in vivo anti-tumor activity in OVCAR-3 xenograft model. The discovery of 13f revealed a more active chemical entity of the HATs domain of p300/CBP and provided a novel idea for the application of epigenetic inhibitors in the treatment of ovarian cancer. [ABSTRACT FROM AUTHOR]

Neural Architecture Search (NAS) methods are widely employed to address the time-consuming and costly challenges associated with manual operation and design of deep convolutional neural networks (DCNNs). Nonetheless, prevailing methods still encounter several pressing obstacles, including limited network architecture design, excessively lengthy search periods, and insufficient utilization of the search space. In light of these concerns, this study proposes an optimization strategy for residual networks that leverages an enhanced Particle swarm optimization algorithm. Primarily, low-complexity residual architecture block is employed as the foundational unit for architecture exploration, facilitating a more diverse investigation into network architectures while minimizing parameters. Additionally, we employ a depth initialization strategy to confine the search space within a reasonable range, thereby mitigating unnecessary particle exploration. Lastly, we present a novel approach for computing particle differences and updating velocity mechanisms to enhance the exploration of updated trajectories. This method significantly contributes to the improved utilization of the search space and the augmentation of particle diversity. Moreover, we constructed a crime-dataset comprising 13 classes to assess the effectiveness of the proposed algorithm. Experimental results demonstrate that our algorithm can design lightweight networks with superior classification performance on both benchmark datasets and the crime-dataset. [ABSTRACT FROM AUTHOR]

Nacre like graphene based composite films (GCFs) have been developed over the last decade. Although the progress in ultrastrong GCFs is remarkable, fabricating GCFs with excellent integrated properties of strength, flexibility, and toughness still remains a challenge. Here, a ternary GCF of reduced graphene oxide (rGO), polydopamine (PDA) and multi-walled carbon nanotubes (MWCNTs) was developed. The MWCNTs were modified by PDA to form PDA@MWCNTs which were then compounded with GO to fabricate the rGO-PDA@MWCNT films by evaporation-induced assembly and chemical reduction. The preparation was simple, cheap and environmentally-friendly. The PDA increased the dispersion of MWCNTs in the rGO-PDA@MWCNT films, crosslinked the rGO sheets, and enhanced the entanglement of MWCNTs and the interactions between MWCNTs and rGO sheets. Benefiting from the synergistic effect of PDA and MWCNTs, the rGO-PDA@MWCNT films exhibited excellent integrated properties by increasing strength, tensile fracture strain and toughness to 579 ± 35 MPa, 12.03 ± 0.56%, and 34.03 ± 1.60 MJ/m3, which were 2.26, 2.02 and 4.94 times that of pure rGO films, respectively. The rGO-PDA@MWCNT films also exhibited a high conductivity of 612 ± 68 S/cm and excellent structural stability under extreme environments. These excellent integrated properties of the rGO-PDA@MWCNT films will promote their applications in aerospace engineering, flexible electrical devices and artificial muscle. Image 1 • A ternary nacre like graphene based composite film (GCF) is developed by a simple, cheap and green method. • This GCF is synergistically toughened and reinforced by polydopamine wrapped carbon nanotubes. • This GCF exhibits good mechanical properties, electrical conductivity and structural stability under extreme environments. [ABSTRACT FROM AUTHOR]

Development of high performance polyether-ether-ketone (PEEK)) composites is of great significance for aerospace engineering application. The fused filament fabrication (FFF)-based 3D printing process is promising for manufacturing high performance short carbon fiber (SCF) reinforced PEEK composites. However, a comprehensive investigation of the combined effects of SCF content and annealing condition has not been conducted on the mechanical performances of SCF/PEEK composites; also, elevated temperature mechanical properties that are required for aerospace engineering have not been reported for 3D printed SCF/PEEK composites. This work develops a creative strategy that combines the optimal short carbon fiber (SCF) content with annealing treatment to achieve significant improvements in the mechanical properties both at room and elevated temperatures of FFF-3D printed SCF/PEEK composites. First, the influences of SCF content on the mechanical performances of 3D printed SCF/PEEK composites are studied to determine the optimal fiber content (5 wt%) for the best mechanical performances of non-annealed PEEK composites. Then, the effects of SCF content and annealing on room temperature and elevated temperature mechanical performances are examined for both PEEK and the 5 wt% SCF/PEEK composite. It is elucidated that the increase of interface strength and crystalline region as well as the reduction of residual stress by annealing are all responsible for the observed improvements in mechanical performances. This work presents a comprehensive study of 3D printed short fiber reinforced thermoplastics and serves as a valuable reference for the 3D printing of high-performance thermoplastic composites. [ABSTRACT FROM AUTHOR]

The rapid urbanization of China related to fine particulate (PM 2.5) emissions has attracted global attention, and exploring how urban forms affect PM 2.5 concentration has become a hot topic for sustainable development of cites. Here, we combined multisource data and econometric models to quantify the relationships between urban forms and PM 2.5 concentration in 250 Chinese cities, with explicit consideration of urban area size, population size, economic structure, and geographical location. The results showed that urban expansion had significant positive impacts on PM 2.5 concentration in medium-sized (50–150 km2) and very large-sized cities (>250 km2). Urban form compactness tended to be beneficial for the reduction in PM 2.5 concentration in large cities (5 million < population < 10 million). The urban form-PM 2.5 concentration relationships were also dominated by economic structure change. That is, urban form irregularity/compactness played an important role in PM 2.5 concentration within second cities (the proportion of second industry > 50%) and other cities compared to third cities (the proportion of third industry > 50%). Moreover, a compact and connected urban form was found to be more beneficial for reducing PM 2.5 emissions in the northern region than in other regions. This study illustrates that the urban form-PM 2.5 concentration relationships are sensitive to the differences in urban types and suggests that flexible urban planning strategies can actually help to reduce PM 2.5 concentration in Chinese cities. [ABSTRACT FROM AUTHOR]

Electronic components are becoming miniaturized and complicated, leading to increased levels of heat generation. Developing lightweight and ultrathin materials with superior thermal conductivity is urgently required to improve thermal management. Here, graphene films (GFs) were fabricated by vacuum filtration of graphene oxide-polydopamine (GO-PDA) solutions and graphitization of the obtained GO-PDA films. In the GO-PDA films, the GO sheets were crosslinked and reduced by PDA. Graphitization transformed PDA into carbonized PDA nanoparticles (CPDA NPs) with the crystal structure of nanosized graphites. The graphene sheets of GFs were covalently crosslinked and reinforced by the CPDA NPs to acquire a dense and flat structure with good sheet alignment and high density, rendering CPDA NP-reinforced GFs with a tensile strength of 47 MPa, an electrical conductivity of 0.89 × 106 S m−1, and a high thermal conductivity of 1584 W m−1 K−1. In comparison, the GFs without the PDA only have a tensile strength of 6.4 MPa, an electrical conductivity of 0.30 × 106 S m−1, and a thermal conductivity of 244 W m−1 K−1. Image 1 [ABSTRACT FROM AUTHOR]

Deep reinforcement learning (DRL) algorithms with experience replays have been used to solve many sequential learning problems. However, in practice, DRL algorithms still suffer from the data inefficiency problem, which limits their applicability in many scenarios, and renders them inefficient in solving real-world problems. To improve the data efficiency of DRL, in this paper, a new multi-step method is proposed. Unlike traditional algorithms, the proposed method uses a new return function, which alters the discount of future rewards while decreasing the impact of the immediate reward when selecting the current state action. This approach has the potential to improve the efficiency of reward data. By combining the proposed method with classic DRL algorithms, deep Q-networks (DQN) and double deep Q-networks (DDQN), two novel algorithms are proposed for improving the efficiency of learning from experience replay. The performance of the proposed algorithms, expected n-step DQN (EnDQN) and expected n-step DDQN (EnDDQN), are validated using two simulation environments, CartPole and DeepTraffic. The experimental results demonstrate that the proposed multi-step methods greatly improve the data efficiency of DRL agents while further improving the performance of existing classic DRL algorithms when incorporated into their training. • A novel multi-step Q-learning method is proposed to improve data efficiency for DRL. • The proposed multi-step Q-learning method is derived by adopting a new return function. • The new return function alters the discount of future rewards and loosens the impact of the immediate reward. • Experimental-results shows the proposed methods can improve the data efficiency of DRL agents. [ABSTRACT FROM AUTHOR]

Highlights • Bilinear isometric mapping is proposed to extract embedding of Riemannian manifold. • The proposed method can maximize the preservation of Riemannian geodesic distance. • A supervised classification algorithm based on extracted embedding is proposed. • High performances of proposed methods are supported by the experimental results. Abstract Because covariance features with the form of symmetric positive definite matrices lie on Riemannian manifold, classification on Riemannian manifold could possess high performance in many applications. Unfortunately, the applicability of classification methods developed on Riemannian manifold is limited by their huge computational complexities, particularly with the feature data on high-dimensional Riemannian manifold. To alleviate the problem of computational cost, in this paper, a simple yet efficient dimensionality reduction algorithm, bilinear isometric Riemannian embedding, is derived to construct a low-dimensional embedding from high-dimensional Riemannian manifold. To this end, we model the bilinear isometric mapping to identify a low-dimensional embedding that maximizes the preservation of Riemannian geodesic distance. A supervised classification method, embedding discriminant analysis, is then proposed based on the low-dimensional embedding. Experimental results on image and electroencephalogram reveal that the proposed algorithms can efficiently extract the distance-preserving embedding and obtain higher classification performance. [ABSTRACT FROM AUTHOR]

Electromagnetic (E/MEG) brain source imaging involves challenging problems that make it particularly difficult to estimate both the locations and extents of extended sources. In this study, we propose a new method called V ariation S parse S ource I maging based on C onditional M ean of the posterior (VSSI-CM), which is built upon a Bayesian framework, to reconstruct extended E/MEG generators. Based on the proposed framework, VSSI-CM can employ various spatial priors (e.g., the Laplace prior) to explore sparseness of current sources in transform domains (e.g., the variation transform in this study). Considering the complexity of posterior density in the estimated sources, we propose using the posterior mean instead of the typical maximum a posterior (MAP) estimate as a more accurate inverse solution. The posterior mean is obtained by fitting an approximated Gaussian distribution to the intractable true posterior distribution. An efficient double-loop algorithm is also proposed using convex analysis skills. Validation using synthetic and human experimental data sets indicates that VSSI-CM outperforms the well-studied L 2 -norm methods (i.e., sLORETA and dSPM) and the sparse constrained methods that explore sparseness in the original source domain. The estimates from VSSI-CM are also more accurate than that from MAP. [ABSTRACT FROM AUTHOR]

The soft error rate (SER) due to heavy-ion irradiation of a clock tree is investigated in this paper. A method for clock tree SER prediction is developed, which employs a dedicated soft error analysis tool to characterize the single-event transient (SET) sensitivities of clock inverters and other commercial tools to calculate the SER through fault-injection simulations. A test circuit including a flip-flop chain and clock tree in a 65 nm CMOS technology is developed through the automatic ASIC design flow. This circuit is analyzed with the developed method to calculate its clock tree SER. In addition, this circuit is implemented in a 65 nm test chip and irradiated by heavy ions to measure its SER resulting from the SETs in the clock tree. The experimental and calculation results of this case study present good correlation, which verifies the effectiveness of the developed method. [ABSTRACT FROM AUTHOR]

Most of the existing clustering methods have difficulty in processing complex nonlinear data sets. To remedy this deficiency, in this paper, a novel data model termed Hybrid K-Nearest-Neighbor (HKNN) graph, which combines the advantages of mutual k-nearest-neighbor graph and k-nearest-neighbor graph, is proposed to represent the nonlinear data sets. Moreover, a Clustering method based on the HKNN graph (CHKNN) is proposed. The CHKNN first generates several tight and small subclusters, then merges these subclusters by exploiting the connectivity among them. In order to select the optimal parameters for CHKNN, we further propose an internal validity index termed K-Nearest-Neighbor Index (KNNI), which can also be used to evaluate the validity of nonlinear clustering results by varying a control parameter. Experimental results on synthetic and real-world data sets, as well as that on the video clustering, have demonstrated the significant improvement on performance over existing nonlinear clustering methods and internal validity indices. [ABSTRACT FROM AUTHOR]

Detecting P300 signals from electroencephalography (EEG) is the key to establishing a P300 speller, which is a type of brain–computer interface (BCI) system based on the oddball paradigm that allows users to type messages simply by controlling eye-gazes. The convolutional neural network (CNN) is an approach that has achieved good P300 detection performances. However, the standard CNN may be prone to overfitting and the convergence may be slow. To address these issues, we develop a novel CNN, termed BN 3 , for detecting P300 signals, where Batch Normalization is introduced in the input and convolutional layers to alleviate over-fitting, and the rectified linear unit (ReLU) is employed in the convolutional layers to accelerate training. Since our model is fully data-driven, it is capable of automatically capturing the discriminative spatio-temporal features of the P300 signal. The results obtained on previous BCI competition P300 data sets show that BN 3 both achieves the state-of-the-art character recognition performance and that it outperforms existing detection approaches with small flashing epoch numbers. BN 3 can be used to improve the character recognition performance in P300 speller systems. [ABSTRACT FROM AUTHOR]