This paper comprehensively investigates the purge mechanism of proton exchange membrane fuel cells during shutdown process, which qualitatively examines effect parameters (including current density, stoichiometric ratio, and relative humidity) on water content variation, further quantitatively remaining post-purge. In contrast to previous studies, this offers a novel perspective analyzing process conducts thorough examination residual content. study presents transient, isothermal, two-phase...

Online reviews allow consumers to provide detailed feedback on various aspects of items. Existing methods utilize these model users' fine-grained preferences for specific item features through graph neural networks. We argue that the performance items different is important making precise recommendations, which has not been taken into account by existing approaches, due lack data. In this paper, we propose an aspect performance-aware hypergraph network (APH) review-based recommendation,...

Knowledge graphs (KGs) model entities or concepts and their relations in a structural manner. The incompleteness has turned out to be the main challenge that hinders application of KG. Recently, reinforcement learning (RL) been recognized as an effective method deal with such challenge, which models research tasks into sequence decision problem without labels. Although increasing number studies investigate analyze knowledge using learning, there lacks systematic literature review...

Hyperspectral image (HSI) classification is one of the most challenging problems in understanding HSI. Convolutional neural network(CNN), with strong ability to extract features using hidden layers network, has been introduced solve this problem. However, several fully connected are always appended at end CNN, which dramatically reduced efficiency space utilization and make algorithm hard converge. Recently, a new network architecture called capsule (CapsNet) was presented improve CNN. It...

Currently, the carbon emissions of building energy consumption account for a significant portion all emissions. How to reduce achieve neutrality is an important current research direction. Therefore this builds predictive algorithm model analyzing data meteorological buildings using DeST platform saving and emission reduction neutrality. The new uses Internet Things cloud technology build simulation platform, support vector machine in analysis vectorize data, which can normalization...

Summary Industrial Wireless Sensor Networks (IWSNs) have gained significant popularity for their ability to improve plant productivity and production efficiency through self‐organization rapid deployment. However, the challenge of achieving reliable sustainable data transmission remains due large amount heterogeneous generated by large‐scale IWSNs. In this paper, we present a systematic approach that addresses focusing on clustering strategies, optimal cluster head selection, routing design....

Zero-knowledge (ZK) proof systems have emerged as a promising solution for building security-sensitive applications. However, bugs in ZK applications are extremely difficult to detect and can allow malicious party silently exploit the system without leaving any observable trace. This paper presents Coda, novel statically-typed language zero-knowledge Critically, Coda makes it possible formally specify statically check properties of application through rich refinement type system. One key...

Music content has recently been identified as useful information to promote the performance of music recommendations. Existing studies usually feed low-level audio features, such Mel-frequency cepstral coefficients, into deep learning models for However, features cannot well characterize audios, which often contain multiple sound sources. In this paper, we propose model and fuse chord, melody, rhythm meaningfully so improve recommendation. Specially, use two user-based attention mechanisms...

Massive open online courses (MOOCs) have been widely used and many institutions invested considerable effort in developing, promoting delivering such courses. Online educational pattern, both inside outside of the campus community, become more popular. In recent years, a new research agenda has emerged, focusing on predicting explaining dropouts MOOCs with low completion rates. However, due to different problem specifications evaluation metrics, performing learning analytics (LA) education...

Entity alignment (EA) aims to find equivalent entities in knowledge graphs (KGs) from multiple data sources and is a crucial step integrating KGs. Recent studies learn the similarity of entity embeddings by aggregating neighboring entities. However, these methods solely compare do not incorporate connected relation between an its neighbors. In this paper, we propose novel Relation joint Interaction Learning (ERIL) approach, which effectively captures interaction relations, enhancing...

Because the current program design curriculum knowledge system is insufficient, this paper puts forward double clue and redesigns teaching content, author suggests that there should be two main clues which are data representation realization in whole process, cross walk. This pays great attention to students' programming thinking ability, makes students not only gain language get ability. So, it raise from skill training. With its "fast slow binding, seizing big freeing small" strategy,...

At present, the speed of communications on PCI is limited by bus, number computer's memory slots and capacity significantly increased, management becomes advanced. So, in this paper, direct communication method, abbreviated as DMC, forwarded. The NIC (the Network Interface Card) based DMC inserted into a slot, data written while it's NIC, user can accomplished point to between computers through accessing memory. isn't copy omitted. author also applied high fibre channel switch network, has...

Hyperspectral imaging makes it possible to obtain object information with fine spectral resolution as well spatial resolution, which is beneficial a wide array of applications. However, there high correlation among the bands in hyperspectral image (HSI). Band selection (BS), selecting only some representative describe original image, an appropriate approach tackle this problem. In study, authors propose efficient greedy‐based unsupervised BS method, namely maximum simplex volume by...