To address challenges in stock price prediction including data nonlinearity and anomalies, we propose a hybrid CNN-BiGRU-AM framework integrated with deep learning-based anomaly detection. First, an detection module identifies irregularities data. The CNN component then extracts local features while filtering anomalous information, followed by nonlinear pattern modeling through BiGRU attention mechanisms. Final predictions undergo secondary screening to ensure reliability. Experimental...

In this study, we propose a novel method for identifying lithology using an attention mechanism-enhanced graph convolutional neural network (AGCN). The aim of is to address the limitations traditional approaches that evaluate unbalanced by improving identification thin layers and small samples, while providing reliable data support reservoir evaluation. To achieve goal, begin Principal Component Analysis (PCA) with maximum minimum distance clustering (Max-min-distance) correct logging...

The precise extraction of displacement time series for complex landslides poses significant challenges, and conventional landslide prediction models often overlook the deformation impacts triggers. To address this, we introduce a novel composite model tailored predicting displacement. This employs Variational Mode Decomposition (VMD) to isolate each component, with optimization achieved through groupwise coupling algorithm. Subsequently, Grey correlation analysis (GRA) is applied...

As small commodity features are often few in number and easily occluded by hands, the overall detection accuracy is low, still a great challenge. Therefore, this study, new algorithm for occlusion proposed. Firstly, super-resolution with an outline feature extraction module used to process input video frames restore high-frequency details, such as contours textures of commodities. Next, residual dense networks extraction, network guided extract information under effects attention mechanism....

Recovery of copper and cobalt from smelter slag using reductive-sulfurizing smelting method was performed in this study. The effects reductive agent (coke), sulfurizing (pyrite), modifier (CaO) temperature duration on the extractive efficiencies Cu, Co Fe were discussed. phase compositions microstructure materials, copper-cobalt matte cleaned determined. results showed that contents could decrease averagely to 0.18% 0.071% respectively after cleaning. 91.99% Cu 92.94% less than 38.73%...

The paper currency characteristic data acquisition system is a real-time subsystem of bill acceptor based on ARM and uClinux, but uClinux can not meet the requirement system. After considering hardware system, method for improving ability proposed by reference architecture RTLinux. A control layer introduced into this to interrupts. task put kernel level as high priority layer, regarded low they are scheduled layer. In experiments, improved requirement.

An approach to fault diagnosis for analog circuits with tolerance is presented based on probabilistic neural networks. In order overcome the difficulties in BP network such as slow learning speed convergence and easily falling into local minimum value, introduced tolerance-circuit diagnosis. Fault samples including soft faults hard are generated by Monte Carlo analysis. features extracted using largest deviation path so obtain appropriate training samples. Simulation results show that...

An approach to analog fault diagnosis for circuits with tolerance is presented based on data fusion. The proposed strategy consists of preliminary and decision Preliminary performed separately by using one kind signatures from multiform circuit responses. results are aggregated an evidential reasoning algorithm location accomplished according the fusion regulation. experimental show that can produce a more meaningful accurate because combining evidences effectively has capability diagnose...

Mobility is a common character of the emerging RFID-based internet-of-things. However, most prior RFID anticollision algorithms ignore movement tags, which can degrade identification performance seriously and even result in tag starvation problem. This paper presents novel algorithm named Query Splitting-based Anticollision (QSA) for mobile By designing adaptive query, QSA reduces number collisions efficiently makes it possible to identify multiple tags without rollback. In QSA, we propose...

To reduce the cargo loss rate caused by abnormal consumption behavior in smart retail cabinets, two problems need to be solved. The first is that diversity of consumers leads a actions contained same behavior, which makes accuracy consumer identification low. Second, difference between normal interaction and small, anomalous features are difficult define. Therefore, we propose an detection algorithm with human–object graph convolution confidence-guided enhancement. Aiming solve problem low...

Dynamic visual vending machines are rapidly growing in popularity, offering convenience and speed to customers. However, there is a prevalent issue with consumers damaging goods then returning them the machine, severely affecting business interests. This paper addresses from standpoint of defect detection. Although existing industrial detection algorithms, such as PatchCore, perform well, they face challenges, including handling various orientations, speeds that do not meet real-time...

Abstract In recent years, defending against adversarial examples has gained significant importance, leading to a growing body of research in this area. Among these studies, pre-processing defense approaches have emerged as prominent direction. However, existing example techniques often employ single model counter different types attacks. Such strategy may miss the nuances between attacks, limiting comprehensiveness and effectiveness strategy. To address issue, we propose divide-and-conquer...

Zero-watermarking is an emerging distortion-free copyright protection method for volumetric medical images. However, achieving both robustness against various malicious attacks and distinguishability between individual images remains challenging. In this paper, we propose a novel attack-defending contrastive learning zero-watermarking scheme (ADCL-ZW) to tackle the above challenge using deep learning-based representations. our approach, design data enrichment mechanism enhance watermarking...

The leaching behavior of hemimorphite in neutral solution (NH4 +-Cl--H2O) was investigated at high temperature (T > 100?C) under a range experimental conditions. Thermodynamic calculations indicate that the tendency dehydration silica gel is significantly enhanced with increasing temperature. It shown temperature, ammonium chloride concentration, or L/S ratio increased resulted greater efficiency. following optimized conditions were obtained: stirrer speed 400r/min, NH4Cl concentration...

Scratchpad memory is a popular choice for on-chip storage in real-time embedded systems. The allocation of program elements to scratchpad performed at compile time leading predictable access latencies. Current techniques mainly improve the average-case execution tasks. For hard systems, however, worst case (WCET) more important key metric. In this paper, we propose data that aim minimize tasks WCET. We first analyze main problem optimal process and current works algorithms. Then, based on...

PDF HTML阅读 XML下载 导出引用 引用提醒 基于双向令牌的可扩展及可靠的群组成员管理 DOI: 作者: 作者单位: 作者简介: 通讯作者: 中图分类号: 基金项目: Supported by the National Natural Science Foundation of China under Grant Nos.60503007, 90718034 (国家自然科学基金); Fund for Distinguished Young Scholars No.60425310 (国家杰出青年科学基金); Program New Century Excellent Talents in University Ministry Education No.NCET-06-0686 (新世纪优秀人才支持计划); Changjiang and Innovative Research Team No.IRT-0661 (国家教育部长江学者与创新团队发展计划) Scalable Reliable Group Membership Management with...