Edge-Federated Self-Supervised Communication Optimization Framework Based on Sparsification and Quantization Compression

The federated self-supervised framework is a distributed machine learning method that combines federated learning and self-supervised learning, which can effectively solve the problem of traditional federated learning being difficult to process large-scale unlabeled data. The existing federated self-supervision framework has problems with low communication efficiency and high communication delay between clients and central servers. Therefore, we added edge servers to the federated self-supervision framework to reduce the pressure on the central server caused by frequent communication between both ends. A communication compression scheme using gradient quantization and sparsification was proposed to optimize the communication of the entire framework, and the algorithm of the sparse communication compression module was improved. Experiments have proved that the learning rate changes of the improved sparse communication compression module are smoother and more stable. Our communication compression scheme effectively reduced the overall communication overhead.

Decentralized Semi-Supervised Learning for Stochastic Configuration Networks Based on the Mean Teacher Method

The aim of this paper is to broaden the application of Stochastic Configuration Network (SCN) in the semi-supervised domain by utilizing common unlabeled data in daily life. It can enhance the classification accuracy of decentralized SCN algorithms while effectively protecting user privacy. To this end, we propose a decentralized semi-supervised learning algorithm for SCN, called DMT-SCN, which introduces teacher and student models by combining the idea of consistency regularization to improve the response speed of model iterations. In order to reduce the possible negative impact of unsupervised data on the model, we purposely change the way of adding noise to the unlabeled data. Simulation results show that the algorithm can effectively utilize unlabeled data to improve the classification accuracy of SCN training and is robust under different ground simulation environments.

Distant Supervision方法中对齐数据的聚类去噪

基于聚类和寻找表达语义关系的句子模式,提出一种基于聚类去噪的distant supervision方法,解决了传统distant supervision方法在知识库和文本集对齐阶段会引入噪声数据的问题,得到了一种改进的distant supervision方法.实验结果表明,该方法可有效提高关系抽取任务的准确率.

基于Semi-Supervised LLE的人脸表情识别方法

目的为提取有效的鉴别特征和降低鉴别向量的维数来识别人脸表情图像.方法将流行学习(Manifold learning,ML)和半监督学习(Semi-Supervised learning,SSL)结合起来,利用人脸表情图像数据本身的非线性流形结构信息和部分标签信息来调整点与点之间的距离形成距离矩阵,而后基于被调整的距离矩阵进行线性近邻重建来实现维数约简,提取低维鉴别特征用于人脸表情识别.结果该方法能充分利用数据的结构信息和有限的标签信息,使具有标签信息的同类样本之间的距离最小化,不同类数据之间的距离最大化,进而可以有效地提取数据的低维鉴别子流形,使得分类性能要优于非监督的维数约简方法.结论笔者提出的半监督局部线性嵌入算法能有效地提高人脸表情识别的性能.

Key Points in the Supervision of Road Landscaping Project——A Case Study of the South 3^(rd) Ring Road in Yanta District of Xi'an City,China

Characteristics of road landscaping were analyzed,as a kind of "linear" landscaping,it could be classified into urban street landscaping,urban ring road and expressway landscaping.Differences between road landscaping projects and other garden landscaping projects or construction projects were analyzed.Landscaping project for the South 3rd Ring Road of Xi'an City(in Yanta District)was taken for an example to summarize construction experience,discuss present conditions and features of garden landscaping supervision in China.The artistic and physiological properties of garden landscaping,as well as the supervision during maintenance period were stressed,but it was also pointed out that garden landscaping supervision in China was actually not given sufficient attention.On this basis,key points in the supervision of road landscaping project were proposed:understanding design ideas,selection of landscape plants,quality control of concealed works,prevention and control of diseases and insects,construction safety management and overall image;only by effectively implementing these key points,a high-quality road landscaping project could be realized to improve the image of a city and beautify its environment.

Situation assessment for air combat based on novel semi-supervised naive Bayes

A method is proposed to resolve the typical problem of air combat situation assessment. Taking the one-to-one air combat as an example and on the basis of air combat data recorded by the air combat maneuvering instrument, the problem of air combat situation assessment is equivalent to the situation classification problem of air combat data. The fuzzy C-means clustering algorithm is proposed to cluster the selected air combat sample data and the situation classification of the data is determined by the data correlation analysis in combination with the clustering results and the pilots＇ description of the air combat process. On the basis of semi-supervised naive Bayes classifier, an improved algorithm is proposed based on data classification confidence, through which the situation classification of air combat data is carried out. The simulation results show that the improved algorithm can assess the air combat situation effectively and the improvement of the algorithm can promote the classification performance without significantly affecting the efficiency of the classifier.

Supervised descent method for weld pool boundary extraction during fiber laser welding process

In order to obtain a high-quality weld during the laser welding process, extracting the characteristic parameters of weld pool is an important issue for automated welding. In this paper, the type 304 austenitic stainless steel is welded by a 5 kW high-power fiber laser and a high-speed camera is employed to capture the topside images of weld pools. Then we propose a robust visual-detection approach for the molten pool based on the supervised descent method. It provides an elegant framework for representing the outline of a weld pool and is especially efficient for weld pool detection in the presence of strong uncertainties and disturbances. Finally, welding experimental results verified that the proposed approach can extract the weld pool boundary accurately, which will lay a solid foundation for controlling the weld quality of fiber laser welding process.

Effects of supervised movie appreciation on the improvement of college students’ life meaning sense

The purpose of this study was to explore the effects of supervised movie appreciation on improving the life meaning sense among college students. The intervention combined by “pre-video, post counseling” was conducted on the experimental group, while the control group received no intervention. Results have shown that the scores on the subscales of will to meaning, life purpose, life control, suffer acceptance and on the total scale have improved significantly. No gender difference was found on the intervention effect, and participants receiving intervention maintained higher level on related subscales a week later, indicating that supervised movie appreciation is an effective way to improve the life meaning sense among college students.

Problems of Rural Food Safety and Strategies of Constructing Supervision System

This paper expounds the practical necessity of constructing diversified rural food safety supervision system as follows: it is the necessary requirements of guaranteeing people's health and life safety; it is an important component of governmental function of social management and the logical extension of administrative responsibilities; it is the basis of maintaining order of rural society and constructing harmonious society. The main problems existing in the supervision of rural food safety are analyzed as follows: first, the legislative work of rural food safety lags behind to some extent; second, the supervision of governmental departments on rural food safety is insufficient; third, the industrial supervision mechanism of rural food security is not perfect; fourth, the role of rural social organizations in supervising food safety is limited; fifth, the farmers' awareness of food safety supervision is not strong. Based on these problems, the targeted strategies of constructing diversified rural food safety supervision system are put forward as follows: accelerate the legislation of rural food safety, and ensure that there are laws to go by; give play to the dominant role of government, and strengthen administrative supervision on rural food safety; perfect industrial convention of rural food safety, and improve industrial supervision mechanism; actively support the fostering of social organizations, and give play to the role of supervision of organizations; cultivate correct concept of rights and obligations of farmers, and form awareness of food safety supervision.

Research on the Status Quo and Supervision Mechanism of Food Safety in China

Food safety supervision mechanism is a strong guarantee to promote the smooth implementation of China's food safety laws and regulations,and it is implemented through legal,administrative,economic,moral and other integrated policy instruments,as well as media publicity,quality traceability,network tracking,information disclosure and other non-administrative means. Along with strengthening supervision and control means,the people's food safety in China is safeguarded,and the healthy development of the food industry is promoted.

Renal function and physical fitness after 12-mo supervised training in kidney transplant recipients

AIM To evaluate the effect of a 12-mo supervised aerobic and resistance training, on renal function and exercise capacity compared to usual care recommendations.METHODS Ninety-nine kidney transplant recipients(KTRs) were assigned to interventional exercise(Group A; n = 52) and a usual care cohort(Group B; n = 47). Blood and urine chemistry, exercise capacity, muscular strength, anthropometric measures and health-related quality of life(HRQo L) were assessed at baseline, and after 6 and 12 mo. Group A underwent a supervised training three times per week for 12 mo. Group B received only general recommendations about home-based physical activities.RESULTS Eighty-five KTRs completed the study(Group A, n = 44; Group B, n = 41). After 12 mo, renal function remained stable in both groups. Group A significantly increased maximum workload(+13 W, P = 0.0003), V'O2 peak(+3.1 mL/kg per minute, P = 0.0099), muscular strength in plantar flexor(+12 kg, P = 0.0368), height in the countermovement jump(+1.9 cm, P = 0.0293) and decreased in Body Mass Index(-0.5 kg/m^2, P = 0.0013). HRQo L significantly improved in physical function(P = 0.0019), physical-role limitations(P = 0.0321) and social functioning scales(P = 0.0346). Noimprovements were found in Group B.CONCLUSION Twelve-month of supervised aerobic and resistance training improves the physiological variables related to physical fitness and cardiovascular risks without consequences on renal function. Recommendations alone are not sufficient to induce changes in exercise capacity of KTRs. Our study is an example of collaborative working between transplant centres, sports medicine and exercise facilities.

基于主题模型的中文Distant Supervision噪声标注识别方法

针对Distant Supervision关系抽取方法训练语料存在大量噪声的问题,提出一种基于主题模型的噪声标注识别方法。该方法首先分析了中文Distant Supervision实体关系抽取方法面临的关系句子实例结构复杂的问题,然后利用自定义的模式以及模式聚类实现模式表示与聚合,最后使用主题模型识别噪声标注。实验结果表明,文章方法能有效识别噪声标注,用滤除噪声标注后的数据训练实体关系抽取模型,实验证明经过噪声滤除后实体关系抽取性能得到显著改善。

Labeling Malicious Communication Samples Based on Semi-Supervised Deep Neural Network

The limited labeled sample data in the field of advanced security threats detection seriously restricts the effective development of research work.Learning the sample labels from the labeled and unlabeled data has received a lot of research attention and various universal labeling methods have been proposed.However,the labeling task of malicious communication samples targeted at advanced threats has to face the two practical challenges:the difficulty of extracting effective features in advance and the complexity of the actual sample types.To address these problems,we proposed a sample labeling method for malicious communication based on semi-supervised deep neural network.This method supports continuous learning and optimization feature representation while labeling sample,and can handle uncertain samples that are outside the concerned sample types.According to the experimental results,our proposed deep neural network can automatically learn effective feature representation,and the validity of features is close to or even higher than that of features which extracted based on expert knowledge.Furthermore,our proposed method can achieve the labeling accuracy of 97.64%~98.50%,which is more accurate than the train-then-detect,kNN and LPA methodsin any labeled-sample proportion condition.The problem of insufficient labeled samples in many network attack detecting scenarios,and our proposed work can function as a reference for the sample labeling tasks in the similar real-world scenarios.

Clinical Supervision of Midwifery Students at the University Teaching Hospital School of Nursing and Midwifery in Lusaka, Zambia

Introduction: Clinical supervision plays a significant role in the acquisition of psychomotor skills by student midwives during training leading to the provision of high quality midwifery care. However, the acquisition of the psychomotor skills required for successful practice and learning can only be acquired if student midwives are supported and guided by clinical supervisors in their learning environments. Main Objective: The main objective of the study was to determine factors influencing clinical supervision of Student Midwives in Lusaka Urban District. Methodology: A cross-sectional study design that employed both quantitative and qualitative methods was used. The study population comprised 124 Nurses and Midwives working in Lusaka urban District. Self-administered questionnaires were used to collect data from the Nurses and midwives and a focus group discussion guide was used to guide discussions with the student midwives at Lusaka College of Nursing and Midwifery and a total of three focus group discussions were conducted. Quantitative data were entered into by Statistical Package Social Sciences (SPSS) version 22 for windows. Chi-square test was used to test associations among variables. The confidence interval was set at 95% with a significance level of 0.05. Qualitative data were analysed using content analysis to classify words into themes and categories by looking for repeated ideas or patterns of thoughts. Results: The findings showed that a high percentage of the clinical supervisors (89.5%) had not been trained in clinical supervision and most respondents had moderate knowledge on clinical supervision of students. Majority of the respondents (63%) exhibited inadequate supervision skills. The attitudes of all the supervisors towards supervision were positive. A large percentage of respondents (73%) were competent in clinical supervision. However the learning environment considered to be poor by most respondents (61%). The study revealed a significant association between clinical supervision and the respondents’ knowledge on clinical supervision (p-value of 0.00). Conclusion: The current study showed that clinical supervision was untrained and had inadequate supervision skills and the learning environment was poor. There is need therefore to train clinical supervisors and improve the students’ learning environment in order to enhance teaching and learning. The study offers a valuable insight into the factors influencing students’ midwives learning in clinical learning environment.

Evaluation Nursing Students’ Views of Improved Competence Development after Clinical Supervision: An Educative Approach to the WHO Patient Safety Model

Bachelor of Science Nursing (BSN) students’ education comprises both theories and practical aspects. Access to resources is required for the development of a professional identity, which includes gaining technical knowledge and receiving feedback, guidance as well as social and emotional support from clinical supervisors. The aim of this study was to evaluate BSN students’ views of professional development after clinical supervision (CS) during their undergraduate education. An additional aim was to illuminate how competence development was related to the WHO Patient Safety Educational Model. A cross-sectional study was conducted, in which CS was measured as part of a survey completed by a sample of nursing students after their clinical placement at two time-points, namely 2012 and 2013. Statistical descriptive and inferential analyses were used and differences in the responses between Time 1 and Time 2 compared. The benefit of CS for nursing students’ competence development revealed a positive significant relationship between students’ Interpersonal skills and the factor Improved care/skills. There were differences in terms of variables related to the Importance value of CS and Professional skills. The results can be used to inform undergraduate nursing education leaders, teachers and practice partners on individual, group and organisational level in order to enhance patient safety and highlight the importance of CS for BSN students’ professional development.

The Impact of Government Supervision and Consumer Purchasing Behavior on Food Safety

This paper aims at the theoretical analysis to the impact of government supervision and consumer purchasing behavior on food quality security, so as to look for safety strategies and measures to strengthen and improve the level of food safety in China. Reputation mechanism is introduced and Bayesian approach is based on, in which government supervision as well as consumer purchasing behavior is taken as crucial factors to impact on the food quality security. As to the proposed quantitative indicators, government supervision includes exposure rate, fine and etc.;at the same time, consumer purchasing behavior includes consumer’s WTP for security food and consumer expectations to food safety. Taking China’s dairy industry as an example, it makes simulation by Netlog. The results show that consumer purchasing behavior alone has little effect on the dairy companies’ decision-making to be honest or counterfeiting enterprises. However, combination government supervision with purchasing behavior has great impact, and plays very good effects on food safety.

Semi-supervised LIBS quantitative analysis method based on co-training regression model with selection of effective unlabeled samples

The accuracy of laser-induced breakdown spectroscopy(LIBS) quantitative method is greatly dependent on the amount of certified standard samples used for training. However, in practical applications, only limited standard samples with labeled certified concentrations are available. A novel semi-supervised LIBS quantitative analysis method is proposed, based on co-training regression model with selection of effective unlabeled samples. The main idea of the proposed method is to obtain better regression performance by adding effective unlabeled samples in semisupervised learning. First, effective unlabeled samples are selected according to the testing samples by Euclidean metric. Two original regression models based on least squares support vector machine with different parameters are trained by the labeled samples separately, and then the effective unlabeled samples predicted by the two models are used to enlarge the training dataset based on labeling confidence estimation. The final predictions of the proposed method on the testing samples will be determined by weighted combinations of the predictions of two updated regression models. Chromium concentration analysis experiments of 23 certified standard high-alloy steel samples were carried out, in which 5 samples with labeled concentrations and 11 unlabeled samples were used to train the regression models and the remaining 7 samples were used for testing. With the numbers of effective unlabeled samples increasing, the root mean square error of the proposed method went down from 1.80% to 0.84% and the relative prediction error was reduced from 9.15% to 4.04%.

Unsupervised Quick Reduct Algorithm Using Rough Set Theory

Feature selection （FS） is a process to select features which are more informative. It is one of the important steps in knowledge discovery. The problem is that not all features are important. Some of the features may be redundant, and others may be irrelevant and noisy. The conventional supervised FS methods evaluate various feature subsets using an evaluation function or metric to select only those features which are related to the decision classes of the data under consideration. However, for many data mining applications, decision class labels are often unknown or incomplete, thus indicating the significance of unsupervised feature selection. However, in unsupervised learning, decision class labels are not provided. In this paper, we propose a new unsupervised quick reduct （QR） algorithm using rough set theory. The quality of the reduced data is measured by the classification performance and it is evaluated using WEKA classifier tool. The method is compared with existing supervised methods and the result demonstrates the efficiency of the proposed algorithm.