Irony is nowadays a pervasive phenomenon in social networks. The multimodal functionalities of these platforms (i.e., the possibility to attach audio, video, and images to textual information) are increasingly leading...
详细信息
How to quickly and accurately detect the surface defects of objects has always been the focus of computer vision research. In this paper, a defect detection based on AI method for anode copper plate is proposed. First...
详细信息
Cryptography is used by all organizations to protect the data files and ensures confidentiality mainly at the time of sharing and storing in the cloud data storage. The cloud service providers use a wide range of tool...
详细信息
China is a major industrial country in the world. In various construction environments, the falling of construction materials and collisions on construction sites are the main causes of casualties. Accidents caused by...
详细信息
Long-term forecasting is widely used in meteorology, hydrology, and finance. However, non-stationary time series make it hard to make accurate long-term predictions because of their complicated multi-period local-glob...
ISBN:
(纸本)9798350337020
Long-term forecasting is widely used in meteorology, hydrology, and finance. However, non-stationary time series make it hard to make accurate long-term predictions because of their complicated multi-period local-global temporal dynamic patterns. Currently, state-of-the-art methods use transformers or temporal convolutions to obtain global and local temporal dynamic patterns. Nevertheless, the former suffers from the computational complexity of self-attention mechanisms despite having a global temporal receptive field. Despite being able to catch local temporal patterns, the latter requires additional layers to capture global temporal patterns. Moreover, the present research disregards integrating multi-period patterns into longterm forecasting. In this paper, we propose MLGNet to tackle the mentioned challenges, which integrates local and global temporal dynamic patterns with multiple periods for longterm forecasting. In particular, we suggest using the maximal overlap discrete wavelet transform (MODWT) as a multi-period decoupling method to decompose non-stationary time series and apply it for the first time to long-term forecasting. In addition, we suggest a multi-scale encoder-decoder framework to capture and fuse local-global temporal dynamic patterns in each decomposed period. Inception dilated causal convolutions-based encoder and a lightweight MLP-based decoder in the framework capture local and global temporal dynamic patterns in series while avoiding the high computational complexity of self-attention mechanisms. Lastly, we suggest time-separable convolutions for aggregating information on temporal dynamic patterns among multiple periods. The above method helps MLGNet better balance the representation ability of time series in 1D and 2D space. Evaluation of five benchmark datasets shows that MLGNet outperforms traditional and state-of-the-art methods, with relative improvements of 13.8 % and 21.9% for multivariate and univariate long-term forecasting, resp
The seamless integration of intelligent Internet of Things devices with conventional wireless sensor networks has revolutionized data communication for different applications,such as remote health monitoring,industria...
详细信息
The seamless integration of intelligent Internet of Things devices with conventional wireless sensor networks has revolutionized data communication for different applications,such as remote health monitoring,industrial monitoring,transportation,and smart *** and reliable data routing is one of the major challenges in the Internet of Things network due to the heterogeneity of *** paper presents a traffic-aware,cluster-based,and energy-efficient routing protocol that employs traffic-aware and cluster-based techniques to improve the data delivery in such *** proposed protocol divides the network into clusters where optimal cluster heads are selected among super and normal nodes based on their residual *** protocol considers multi-criteria attributes,i.e.,energy,traffic load,and distance parameters to select the next hop for data delivery towards the base *** performance of the proposed protocol is evaluated through the network simulator *** different traffic rates,number of nodes,and different packet sizes,the proposed protocol outperformed LoRaWAN in terms of end-to-end packet delivery ratio,energy consumption,end-to-end delay,and network *** 100 nodes,the proposed protocol achieved a 13%improvement in packet delivery ratio,10 ms improvement in delay,and 10 mJ improvement in average energy consumption over LoRaWAN.
Components are the fundamental building blocks of Android applications. Different functional modules represented by components often rely on inter-component communication mechanisms to achieve cross-module data transf...
详细信息
ISBN:
(纸本)9798350329964
Components are the fundamental building blocks of Android applications. Different functional modules represented by components often rely on inter-component communication mechanisms to achieve cross-module data transfer and method invocation. It is necessary to conduct robustness testing on components to prevent component launching crashes and privacy leaks caused by unexpected input parameters. However, as the complexity of the input parameter structure and the diversity of possible inputs, developers may overlook specific inputs that result in exceptions. At the same time, the vast input space also brings challenges to efficient component testing. In this paper, we designed an automated test generation and execution tool for Android application components named ICTDroid, which combines static parameter extraction and adaptive-strength combinatorial testing generation to detect bugs with a compact test suite. Experiments have shown that the tool triggers 205 unique exceptions in 30 open-source applications with 1,919 test cases in 83 minutes, where the developers have confirmed six defects in three issues we reported.
WebAssembly is an emerging secure instruction set architecture widely used in embedded applications. Despite its success, bugs in WebAssembly virtual machines can compromise safety. This paper presents the first empir...
详细信息
Given the unpredictable nature of the estimation model’s inputs, reliably predicting test efforts with Machine Learning (ML) methods is a difficult challenge. More data simply helps to develop more precise estimates ...
详细信息
Over the past few years afterward the birth of ResNet, skip connection has become the defacto standard for the design of modern architectures due to its widespread adoption, easy optimization, and proven performance. ...
详细信息
暂无评论