Optimized operation scheme of flash-memory-based neural network online training with ultra-high endurance

With the rapid development of machine learning,the demand for high-efficient computing becomes more and more urgent.To break the bottleneck of the traditional Von Neumann architecture,computing-in-memory(CIM)has attracted increasing attention in recent years.In this work,to provide a feasible CIM solution for the large-scale neural networks(NN)requiring continuous weight updating in online training,a flash-based computing-in-memory with high endurance(10^(9) cycles)and ultrafast programming speed is investigated.On the one hand,the proposed programming scheme of channel hot electron injection(CHEI)and hot hole injection(HHI)demonstrate high linearity,symmetric potentiation,and a depression process,which help to improve the training speed and accuracy.On the other hand,the low-damage programming scheme and memory window(MW)optimizations can suppress cell degradation effectively with improved computing accuracy.Even after 109 cycles,the leakage current(I_(off))of cells remains sub-10pA,ensuring the large-scale computing ability of memory.Further characterizations are done on read disturb to demonstrate its robust reliabilities.By processing CIFAR-10 tasks,it is evident that~90%accuracy can be achieved after 109 cycles in both ResNet50 and VGG16 NN.Our results suggest that flash-based CIM has great potential to overcome the limitations of traditional Von Neumann architectures and enable high-performance NN online training,which pave the way for further development of artificial intelligence(AI)accelerators.

Application Effect of Remote Ultrasound Teaching in Standardized Residency Training

Objective:To explore the application effect of remote ultrasound teaching in the standardized training of residents.Methods:42 students who participated in the standardized residency training in the Department of Ultrasonography of our hospital from August 2022 to August 2023 were selected and divided into the control group(n=21)and the observation group(n=21)by using the random number table method.The control group was taught routinely,and the observation group was taught with remote ultrasound on the basis of the control group.The general data,teaching effect,ultrasound diagnostic compliance rate,and teaching satisfaction of the participants in the two groups were observed.Results:The baseline data of the two groups were not statistically significant(P>0.05);the theoretical and practical assessment scores of the observation group were significantly better than those of the control group(t=2.491,t=2.434,P=0.05);the ultrasound diagnostic compliance rate of the participants in the observation group was significantly higher than that of the control group(78.33%)(χ2=33.574,P=0.000<0.001);the overall satisfaction rate of students in the observation group(20/95.24%)was significantly higher than that of the control group(14/66.67%)(χ2=3.860,P=0.049<0.05).Conclusion:In standardized residency training,remote ultrasound teaching can effectively improve the comprehensive ability of students,enhance diagnostic accuracy,and improve students’teaching satisfaction.

Remote nursing training model combined with proceduralization in the intensive care unit dealing with patients with COVID-19

The shortage of personal protective equipment and lack of proper nursing training have been endangering health care workers dealing with coronavirus disease 2019(COVID-19).In our treatment center,the implementation of a holistic care model of time-sharing management for severe and critical COVID-19 patients has further aggravated the shortage of intensive care unit(ICU)professional nurses.Therefore,we developed a short-term specialized and targeted nursing training program to help ICU nurses to cope with stress and become more efficient,thus reducing the number of nurses required in the ICU.In order to avoid possible human-to-human spread,small teaching classes and remote training were applied.The procedural training mode included four steps:preparation,plan,implementation,and evaluation.An evaluation was conducted throughout the process of nursing training.In this study,we documented and shared experiences in transitioning from traditional face-to-face programs to remote combined with proceduralization nursing training mode from our daily work experiences during the COVID-19 pandemic,which has shown to be helpful for nurses working in the ICU.

Effective distributed convolutional neural network architecture for remote sensing images target classification with a pre-training approach

How to recognize targets with similar appearances from remote sensing images(RSIs) effectively and efficiently has become a big challenge. Recently, convolutional neural network(CNN) is preferred in the target classification due to the powerful feature representation ability and better performance. However,the training and testing of CNN mainly rely on single machine.Single machine has its natural limitation and bottleneck in processing RSIs due to limited hardware resources and huge time consuming. Besides, overfitting is a challenge for the CNN model due to the unbalance between RSIs data and the model structure.When a model is complex or the training data is relatively small,overfitting occurs and leads to a poor predictive performance. To address these problems, a distributed CNN architecture for RSIs target classification is proposed, which dramatically increases the training speed of CNN and system scalability. It improves the storage ability and processing efficiency of RSIs. Furthermore,Bayesian regularization approach is utilized in order to initialize the weights of the CNN extractor, which increases the robustness and flexibility of the CNN model. It helps prevent the overfitting and avoid the local optima caused by limited RSI training images or the inappropriate CNN structure. In addition, considering the efficiency of the Na¨?ve Bayes classifier, a distributed Na¨?ve Bayes classifier is designed to reduce the training cost. Compared with other algorithms, the proposed system and method perform the best and increase the recognition accuracy. The results show that the distributed system framework and the proposed algorithms are suitable for RSIs target classification tasks.

An Effective Online Collaborative Training in Developing Listening Comprehension Skills

The COVID-19 outbreak severely affected formal face-to-face classroom teaching and learning.ICT-based online education and training can be a useful measure during the pandemic.In the Pakistani educational context,the use of ICT-based online training is generally sporadic and often unavailable,especially for developing English-language instructors’listening comprehension skills.The major factors affecting availability include insufficient IT resources and infrastructure,a lack of proper online training for speech and listening,instructors with inadequate academic backgrounds,and an unfavorable environment for ICT-based training for listening comprehension.This study evaluated the effectiveness of ICT-based training for developing secondary-level English-language instructors’listening comprehension skills.To this end,collaborative online training was undertaken using random sampling.Specifically,60 private-school instructors in Chakwal District,Pakistan,were randomly selected to receive online-listening training sessions using English dialogs.The experimental group achieved significant scores in the posttest analysis.Specifically,there were substantial improvements in the participants’listening skills via online training.Given the unavailability of face-to-face learning during COVID-19,this study recommends using ICT-based online training to enhance listening comprehension skills.Education policymakers should revise curricula based on online teaching methods and modules.

Impacts of Online Remote Education on the Learning Process among Nursing Students

The period of existence and spread of Corona virus has led to the use of all means of remote education as an urgent necessity for all educational facilities, especially universities. Aim: Therefore, it was necessary to study the impacts of online remote education on the learning process among nursing students through studying of two courses;health information management at 6th semester and gerontology nursing course at 4th semester. Tool of Data Collection: A modified questionnaire comprised of forty statements was used through paper-based survey and online survey. Sample: A total samples (224) of nursing students were participated in the survey who enrolled in 2019 and 2020 spring semesters. Setting: The field of study was the nursing department of Applied Medical Science at Misr University for Science and Technology. Results: Induced positive impacts of online education on the learning process for nursing students experience were proven as more than half of the students (53.9%) had prior experience on online system use, and more than two thirds (62.5%) were competent in mobile/computer applications. Almost, two thirds (59.3%) agreed about online assessment experience, except that the online exam was anxious, and the time was insufficient to answer all questions. Also, more than two thirds (64.7%) agreed about the learning process of the two nursing courses. Conclusion: The study concluded that there were positive impacts of online education system on the learning process for nursing students except that the students were not able to decide that the remote online education system can replace traditional face-to-face learning as the clinical experience was not evaluated through this study. Recommendation: This study is recommended to be repeated on a large scale of participants to assess the possibility of achieving clinical experience through online remote education if Corona virus still coexists.

Online split-and-merge expec tation-maximization training of Gaussian mixture model and its optimization

This paper presents a new online incremental training algorithm of Gaussian mixture model （GMM）, which aims to perform the expectation-maximization（EM） training incrementally to update GMM model parameters online sample by sample, instead of waiting for a block of data with the sufficient size to start training as in the traditional EM procedure. The proposed method is extended from the split-and-merge EM procedure, so inherently it is also capable escaping from local maxima and reducing the chances of singularities. In the application domain, the algorithm is optimized in the context of speech processing applications. Experiments on the synthetic data show the advantage and efficiency of the new method and the results in a speech processing task also confirm the improvement of system performance.

Airway management training program for nurses via online course in COVID-19 preparedness

BACKGROUND Nursing officers are an integral component of any medical team.They participate in taking care of basic airway management and assist in advanced airway management,specifically amidst the current coronavirus disease 2019(COVID-19)pandemic.AIM To assess the efficacy of a standardized web-based training module for nurses in preparedness to fight against COVID-19.METHODS The training was held in three sessions of 1 h each,consisting of live audio-visual lectures,case scenarios,and skill demonstrations.The sequence of airway equipment,drug preparation,airway examination,and plans of airway management was demonstrated through mannequin-based video-clips.RESULTS Pre-and post-test scores as well as objective structured clinical examination scores were analyzed using Student’s t-test and the Likert scale was used for feedback assessment.It was found that the mean score out of the total score of 20 was 8.47±4.2 in the pre-test,while in the post-test it was 17.4±1.8(P value<0.001).The participants also felt self-reliant in executing the roles of airway assistant(63.3%)and drug assistant(74.3%).Fear of self-infection with COVID-19 was also high,as 66%of participants feared working with the patient’s airway.CONCLUSION Amidst this COVID-19 emergency,when the health care systems are being persistently challenged,training of nursing staff in the safe conduct of airway management can ensure delivery of life-saving treatment.

基于Online-GRU信道预测的星上自适应功率控制方法

针对传统卫星功率控制方法存在资源浪费、时延长的问题,提出一种基于在线-门控循环单元(Online-GRU)信道预测的星上自适应功率控制方法,通过在线训练更新网络参数来解决离线预测算法存在的累积误差的问题。仿真结果表明,提出的在线训练算法比离线算法预测精确度提升了38.30%,相比在线-长短期记忆网络(Online-LSTM)节约了63.21%的训练时间;提出的自适应功率控制方法比固定发射功率的方法节约了55.74%的发射功率;同时,相比基于地面定时反馈信道状态的自适应功率控制方法具备更好的鲁棒性。

Building a Smart Management System for the Field Training Course at the Faculty of Specific Education-Mansoura University

Field training is the backbone of the teacher-preparation process.Its importance stems from the goals that colleges of education aim to achieve,which include bridging the gap between theory and practice and aligning with contemporary educational trends during teacher training.Currently,trainee students attendance in field training is recordedmanually through signatures on attendance sheets.However,thismethod is prone to impersonation,time wastage,and misplacement.Additionally,traditional methods of evaluating trainee students are often susceptible to human errors during the evaluation and scoring processes.Field training also lacks modern technology that the supervisor can use in case of his absence from school to monitor the trainee students’implementation of the required activities and tasks.These shortcomings do not meet the needs of the digital era that universities are currently experiencing.As a result,this paper presents a smart management system for field training based on Internet of Things(IoT)and mobile technology.It includes three subsystems:attendance,monitoring,and evaluation.The attendance subsystem uses an R307 fingerprint sensor to record trainee students’attendance.The Arduino Nano microcontroller transmits attendance data to the proposed Android application via an ESP-12F Wi-Fi module,which then forwards it to the Firebase database for storage.The monitoring subsystem utilizes Global Positioning System(GPS)technology to continually track trainee students’locations,ensuring they remain at the school during training.It also enables remote communication between trainee students and supervisors via audio,video,or text by integrating video call and chat technologies.The evaluation subsystem is based on three items:an online exam,attendance,and implementation of required activities and tasks.Experimental results have demonstrated the accuracy and efficiency of the proposed management system in recording attendance,as well as in monitoring and evaluating trainee students during field traiing.

Challenges to Online Criminal Litigation in the Context of Smart Justice and Responses——Focusing on the Protection of the Right to Defense

Online criminal litigation transcends the constraints of physical time and space and changes the logic and path of trial hearings for some criminal cases with the help of technology.However,the leapfrog shift from the“physical field”to the“virtual field”has brought great challenges to the effective exercise of the defendant's right to defense.Online criminal justice further highlights the imbalance in the relationship between prosecution and defense in the context of smart justice,and proposes a new topic for protecting the human rights of the prosecuted.The introduction of online criminal litigation in judicial practice is intended to achieve justice in a faster and more convenient way.However,the dissipation of the ritualized remote hearings tends to undermine the effectiveness of the defense and impair the defense's ability to cross-examine evidence,while the technically advantageous public authorities can aggravate the barrier to the defense's meeting and reading the case file.The root cause is that technological power instrumentalism overemphasizes pragmatism and the pursuit of truth under the position of authority,thus diluting humanistic care for the subject of litigation.In order to resolve the problem with the quality and effectiveness of the right to defense in remote hearings,it is necessary to transform online criminal litigation from a“practical technical tool”to a“convenient auxiliary method,”and appropriately weigh the limits of pursuing truth against human rights protection in special scenarios.Meanwhile,it is also feasible to provide technical care for the defense and strengthen its ability to cross-examine evidence.Moreover,a covert communication platform should be furnished for the defender's online meeting to actively strengthen the protection of the defendant's right to defense.

Online SVM在实时入侵检测中的应用研究

Online支持向量机作为一种新的分类方法可以在异常入侵检测中提供良好的分类效果。根据Online算法对传统支持向量机、Robust支持向量机和One-class支持向量机进行改进,将改进后的算法与原始算法进行比较,然后使用1999DARPA数据作为评估数据。通过实验和比较发现,改进后的支持向量机可以实现在线训练,而且使用更少的支持向量,训练时间也有效缩短,在噪声数据存在的情况下检测正确率和虚警率比未改进前有一定程度的提升。

通过NFS实现Informix-Online数据库异地备份

介绍了NFS(网络文件系统)的原理和在SCO UNIX中配置NFS的方法,并给出了一个利用NFS对Inform ix-On line数据库进行异地备份的实例。由于数据异地备份是保证数据安全的基本要求,因此,该方法对需要异地备份数据的应用系统极具实用价值。

用于在线产品评论质量分析的Co-training算法

在线评论广泛存在于电子商务网站平台,其中包含着客户对产品的评价及偏好.高效分析在线评论数据并满足客户需求,对许多谋求立足于竞争激烈的国际化市场的企业来说至关重要.但因在线评论的质量不一,使得如何分析在线评论的质量成为一项重要工作.从两个方面提取特征对在线评论进行描述,并构建了一种Co-training算法来判断评论的质量.通过对比实验验证了该算法相对于单一分类算法的优势.

E-commerce development in rural and remote areas of BRICS countries

E-commerce plays an essential role in modern trade today.It is expected that e-commerce volume amounted to 29 trillion USD in the world in 2017,and would grow with the spread of the Internet and information and communication technologies(ICTs).Brazil,Russia,India,China and South Africa(BRICS),together with many others,consider e-commerce a means to facilitate rapid,inclusive and sustainable economic growth,improving the living standards and alleviating poverty.This article examines areas for potential cooperation by BRICS countries in e-commerce development across rural and remote areas to fight poverty.It analyses the current state of e-commerce development in rural and remote areas in each of the BRICS countries,including cases of public and private initiatives to support it.The article also defines the opportunities which e-commerce brings to people living in rural and remote areas.Moreover,it evaluates the existing challenges and risks.The article concludes that despite the rapid e-commerce development in BRICS countries,and significant opportunities created,there are still issues of disproportionate e-commerce in varied regions and the lack of BRICS cooperation in this sphere.Based on a comparative and normative in-depth,systematic analysis,the article develops a set of recommendations for deepening BRICS countries'cooperation in the following areas:infrastructure in rural and remote regions;education;consumer protection;online dispute resolution;coordinated policy in the international scene,including representation of BRICS countries in international indexes,such as the Organization of Economic Co-operation and Development(OECD)Digital Services Trade Restrictiveness Index(STRI).

Online LS-SVM for function estimation and classification

An online algorithm for training LS-SVM (Least Square Support VectorMachines) was proposed for the application of function estimation and classification. Online LS-SVMmeans that LS-SVM can be trained in an incremental way, and can be pruned to get sparseapproximation in a decremental way. When a SV (Support Vector) is added or removed, the onlinealgorithm avoids computing large-scale matrix inverse. Thus the computation cost is reduced. Onlinealgorithm is especially useful to realistic function estimation problem such as systemidentification. The experiments with benchmark function estimation problem and classificationproblem show the validity of this online algorithm.

A RBF classification method of remote sensing image based on genetic algorithm

The remote sensing image classification has stimulated considerable interest as an effective method for better retrieving information from the rapidly increasing large volume, complex and distributed satellite remote imaging data of large scale and cross-time, due to the increase of remote image quantities and image resolutions. In the paper, the genetic algorithms were employed to solve the weighting of the radial basis faction networks in order to improve the precision of remote sensing image classification. The remote sensing image classification was also introduced for the GIS spatial analysis and the spatial online analytical processing (OLAP), and the resulted effectiveness was demonstrated in the analysis of land utilization variation of Daqing city.

利用RemoteApp实现涉密电子信息集中管控系统的安全在线编辑

从涉密电子信息分散存储在用户的电脑、手机或移动存储,到采用涉密电子信息集中管控系统来集中管控涉密电子信息,体现了质的飞跃.然而,涉密电子信息集中管控系统对电子文件的在线编辑必须依赖本地第三方软件,存在失泄密的隐患.提出一种基于RemoteApp的安全在线编辑方法,从根本上巩固了电子信息在编辑环节的安全性.

Smart Photogrammetric and Remote Sensing Image Processing for Very High Resolution Optical Images——Examples from the CRC-AGIP Lab at UNB

This paper introduces some of the image processing techniques developed in the Canada Research Chair in Advanced Geomatics Image Processing Laboratory (CRC-AGIP Lab) and in the Department of Geodesy and Geomatics Engineering (GGE) at the University of New Brunswick (UNB), Canada. The techniques were developed by innovatively/“smartly” utilizing the characteristics of the available very high resolution optical remote sensing images to solve important problems or create new applications in photogrammetry and remote sensing. The techniques to be introduced are: automated image fusion (UNB-PanSharp), satellite image online mapping, street view technology, moving vehicle detection using single set satellite imagery, supervised image segmentation, image matching in smooth areas, and change detection using images from different viewing angles. Because of their broad application potential, some of the techniques have made a global impact, and some have demonstrated the potential for a global impact.