Automated ECG-based arrhythmia detection is critical for early cardiac disease prevention and diagnosis. Recently, deep learning algorithms have been widely applied with great success. However, the lack of labeled ECG data low classification accuracy can a significant impact on overall effectiveness algorithm. In order to better apply methods classification, in this study, feature extraction strategy based generative adversarial network augmentation model fusion are proposed address these...

Safer, more convenient methods for cervical sample collection and storage are necessary to facilitate human papillomavirus (HPV) DNA testing in low-resource settings. Our study aimed evaluate the stability of specimens collected with dry swabs stored dry, compared liquid-based cytology (LBC) samples, as detected by HPV testing. Women abnormal cytological findings or HPV-positive results at colposcopy were recruited from West China Second University Hospital, Sichuan University, between...

Counterfeit products are internationally regarded as “the world’s second greatest public health hazards after drugs”. Counterfeiters produce counterfeit brand clothing and then sell them to consumers through unofficial channels; thus, spend a lot of money without getting the value they deserve. With rise e-shopping, safety security branded supply chains also under threat. apparel manufacturers generate profits while genuine suffer, which ultimately violates interests public. This study...

The automatic diagnosis of various retinal diseases based on fundus images is important in supporting clinical decision-making. Convolutional neural networks (CNNs) have achieved remarkable results such tasks. However, their high expression ability possibly leads to overfitting. Therefore, data augmentation (DA) techniques been proposed prevent overfitting while enriching datasets. Recent CNN architectures with more parameters render traditional DA insufficient. In this study, we a new...

Scientific papers are an important form for researchers to summarize and display their research results. Information mining analysis of scientific can help a comprehensive understanding the subject. Aiming at ignorance contextual semantic information in current topic uncertainty screening rules association evolution research, this paper proposes model based on BERT-LDA model. First, combines learned by BERT with word vectors LDA mine deep topics. Then construct filtering eliminate invalid...

In recent years, the attention of online cross-university courses has been increasing, and students in universities want to increase their knowledge professional skills by taking from different universities, which raises issue course credit verification. past, credits obtained lack endorsement education department, students’ learning process could not be verified. Therefore, one university recognized other universities. The departments some countries regions implement conversion rules...

Inductively coupled plasma (ICP) is a well-known source for atomic spectroscopy, either using optical emission spectroscopy (OES) or mass spectrometry (MS) as detection technique. The high energy and temperatures in an ICP (typically between 6,000 to 8,000 °C) lead efficient complete decomposition of the sample well vaporization, excitation, possibly ionization atoms their gas phase. temperature makes this inherently robust tolerant wide variety types (such solutions containing elevated...

With the rapid development of industrial Internet Things, secure transmission nodes in Things is easily attacked by malicious at all times, thereby causing a great threat to network performance and security Things. To improve success rate data environment, this work proposes scheme based on weighted partition model considering requirements hostile environment. The determines secret distribution model. On basis, authentication for proposed prevent from being tampered nodes. Authority access...

Cui, J., 2018. Dynamic migration algorithm of marine big data in cloud computing environment. In: Liu, Z.L. and Mi, C. (eds.), Advances Sustainable Port Ocean Engineering. Journal Coastal Research, Special Issue No. 83, pp. 706–712. Coconut Creek (Florida), ISSN 0749-0208.Aiming at the shortcomings current based on importance data, like poor load balancing large consumption resources, paper proposes a dynamic for Firstly, characteristics are analyzed. And then Moran's I index is used to...

This paper proposes an image information hiding algorithm which bases on the HVS and MBNS. uses characteristics of Human Visual System (HVS) to embedded more large amounts data according different area achieve degree embedded, pixels in smooth areas corresponding base small, complex great, then secret segment can be into based modular arithmetic. A case study is provided demonstrate fact that this method efficient load ensure lowest rate distortion relative most standard test images.

The training of traditional SVM method requires the solution quadratic programming, and consumes high memory has low speed for large data training.Incremental learning is one meaningful methods to continuously update learning, which keeps previous results, re only additional data, so as form a continuous process.This paper will study support vector machine based on incremental its application in malware detection.The experiments carried out Internet Security Laboratory at Kingsoft...

Abstract This paper investigates the application of exploring spatial clustering techniques in travel recommendation systems to improve accuracy and user satisfaction. A system that can provide personalized information has been designed by analyzing data behavioral patterns travelers. With this approach, experience be optimized considering geographic location, preferences, tourist attraction characteristics. User characteristics are analyzed obtain accurate recommendations. Combining...

Abstract Located in Qingyang, Gansu province, China, the Nanzuo site on Loess Plateau is a large‐scale, high‐ranking central settlement of late Yangshao period with capital city characteristics. We used thermal dilatometer, energy‐dispersive X‐ray fluorescence (EDXRF), optical microscope, scanning electron microscopy spectroscopy (SEM‐EDS), diffraction (XRD) and Raman spectrometer to analyse firing temperature, composition, microstructure phase composition white potteries site. The results...

In the field of software reengineering, many component identification approaches have been proposed for evolving legacy systems into component-based systems. Understanding behaviors various is first important step to meaningfully employ them evolution, therefore we performed an empirical study on technology with considerations their similarity measures, clustering and stopping criteria. We a set evaluation criteria developed tool CIETool automate process evaluation. The experimental results...

One fundamental problem with the kernel Fisher discriminant analysis (KFDA) for fault diagnosis, is singularity of within-class scatter matrix due to small sample size. In this paper, a scatter-difference-based (KSDA) method proposed diagnosis. The can not only produce nonlinear features process data, but also avoid matrix. Experimental results are given show effectiveness new method.

The classic word embedding-based Chinese keyword extraction techniques have a number of drawbacks, including poor segmentation and an inability to recognize ambiguity. An efficient ALBERT-BiGRU-CRF-based model has been presented address the aforementioned issues. first obtains dynamic character vector representation based on ALBERT pre-training model, fuses lexical information by enhancement method, then extracts contextual long-range semantic features using bidirectional gated recurrent...

With the development of machine learning, fair learning has started to receive gradual attention. How mitigate or eliminate possible unfair decision results become a popular research topic in this field. At present, on is still its initial stage. In paper, we analyzed and articles related (January 2011 December 2022) using CiteSpace visualization software, explored collaboration networks (authors, institutions, countries), keyword co-occurrence clustering networks, literature co-citation...