Search

Copyright of Journal of Guangdong University of Technology is the property of Journal of Guangdong University of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Application Research of Computers / Jisuanji Yingyong Yanjiu is the property of Application Research of Computers Edition and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Frontiers of Computer Science & Technology is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Guangdong University of Technology is the property of Journal of Guangdong University of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Computer Engineering & Applications is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Guangdong University of Technology is the property of Journal of Guangdong University of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Guangdong University of Technology is the property of Journal of Guangdong University of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Guangdong University of Technology is the property of Journal of Guangdong University of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Guangdong University of Technology is the property of Journal of Guangdong University of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Application Research of Computers / Jisuanji Yingyong Yanjiu is the property of Application Research of Computers Edition and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Recommendation systems are widely used everywhere and have a great influence on daily life. Aiming to train an ideal recommendation system, a massive of ‘use-item’ interactive pairs should be provided. However, the dataset obtained is usually sparse, which might result in an over-fitting model and be hard to obtain the satisfying performance. In order to address this problem, the cross-domain recommendation is raised. Most of the existing methods on cross-domain recommendation systems borrow the ideas of the conventional unsupervised domain adaptation, which employ the feature alignment or adversarial training methods. Hence they can transfer the domain-invariant interests of users from the source to the target domains, e. g., from the Douban Movies to the Douban Books. However, since the network structures vary with different recommendation platforms, the existing methods on cross-domain recommendation systems straightforwardly extract the domain-invariant representation may entangle the structure information, which may result in the false alignment phenomenon. Furthermore, the previous efforts ignore the noise information behind the graph data, which further degenerate the experimental performance. In order to address the aforementioned problems, this paper brings the causal generation process of graph data into the cross-domain recommendation systems, this paper use the semantic latent variables of each node to calculate the relationships between users and items via disentangling the semantic latent variables, domain latent variables and noise latent variables. Experiments studies testify that the proposed method yields state-of-the-art performance on several cross-domain recommendation system benchmark datasets. [ABSTRACT FROM AUTHOR]

Copyright of Application Research of Computers / Jisuanji Yingyong Yanjiu is the property of Application Research of Computers Edition and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Granger causality discovery algorithm for discrete-time series data has important application value. The existing method smainly use Hawkes processes modeling, which cannot be applied to non-IID data and data with time-skew errors. Therefore, this paper proposed a Granger causality discovery algorithm (PTHP) for topological Hawkes processes integrating a priori constraints. Firstly, it used the constraint-based method to screen a group of causal edges with a high significance level to improve the tolerance of the algorithm to the fault time-skew errors. Then, the edges obtained in the previous step were fused into the topological Hawkes processes as a priori constraints to solve the non-IID problem between sequences. Experiments on simulted data and real-world data show the effectiveness of this method, and it won first place in PCIC 2021 causal inference competition. [ABSTRACT FROM AUTHOR]

Copyright of Application Research of Computers / Jisuanji Yingyong Yanjiu is the property of Application Research of Computers Edition and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In the field of machine learning, graph neural network plays an increasingly important role in works such as social recommendation compared with traditional neural network, but most of the current work use static graph. Aiming at the problem hat the existing static graph neural network method is difficult to consider the dynamic characteristics of social users, this paper proposed a method of social network node classification based on a heterogeneous dynamic graph model by introducing a dynamic graph model. On the basis of dynamic graph modeling, this method realized efficient node classification of dynamic social network through node feature updating mechanism based on point-edge interaction and time series aggregation method based on recurrent neural network. The experimental results on multiple real datasets show that the proposed method has a good effect in dynamic social network data node classification, which has a significant improvement compared with the reference method of the I static graph and dynamic graph. [ABSTRACT FROM AUTHOR]

Copyright of Application Research of Computers / Jisuanji Yingyong Yanjiu is the property of Application Research of Computers Edition and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Computer Engineering & Applications is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Causal discovery from the observational data set has attracted more and more research scholars ' attentions in recent years, where the exogenous variable plays an important role in understanding the causal mechanism. However, most existing causal discovery algorithms assume that the observed variables are the real causal variables, and ignore the effects of measurement errors. This paper proposed a method to estimate the exogenous variables under measurement error model. Specifically, by introducing the state-of-the-art triad constraint, the variable that satisfied triad constraint with all the other related paired variables was the exogenous variable. This method could not only solve the data set with measurement errors, but also handle with the data set without measurement errors. Applied the algorithm to the data generated by real networks, experimental results show that the proposed method is superior to other existing algorithms regardless of whether the variables contain measurement errors. Meanwhile, the experiment of the real mobile-base-station data also verifies the effectiveness of the algorithm. [ABSTRACT FROM AUTHOR]

Existing causal discovery algorithms are mainly based on the observed variables, and cannot be applied to the causal discovery among multiple sets of observed variables. Hence, this paper proposed a multiset canonical correlation variables based causal discovery algorithm. First, it introduced multiset canonical correlation variables to establish a linear non-Gaussian acyclic model for them,and proposed a corresponding objective function. Then,it used the gradient as cent method to solve the objective function,and constructed a causal network over multiset canonical correlation variables. Simulation experiments verify the correctness and effectiveness of the algorithm, and find a number of valuable sets of wireless network performance indicators on the mobile base station dataset. [ABSTRACT FROM AUTHOR]

Time-series causal discovery for high-dimensional networks have been increasingly significant in social media causality. However, the existing algorithms' inability to discover direct causal relations renders the results of causal network inference not so accurate. Hence, this paper proposed a direct causal network discovery algorithm. It considered various factors, including the causal delay, the lag length and the conditional nodes sets, in the time-series causal model to help improve the accuracy of direct causal network inferring. Further, the method solved the difficulty of setting transfer entropy thresholds through permutation test. Experimental results demonstrate that the method outperforms the existing algorithms in causal network inferring and can conspicuously improve the accuracy of direct causal network inference on time series, which is suitable for discovering potential causal networks in social media. [ABSTRACT FROM AUTHOR]

Copyright of Application Research of Computers / Jisuanji Yingyong Yanjiu is the property of Application Research of Computers Edition and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Application Research of Computers / Jisuanji Yingyong Yanjiu is the property of Application Research of Computers Edition and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In the unsupervised domain adaptation area,loss of label information is still an open problem during the alignment of the global distribution, and thus the effect of transfer learning is difficult to guarantee. To alleviate this problem, this paper proposed an algorithm which adopted a distribution alignment strategy based on the intra-class maximum mean discrepancy. This strategy firstly predicted the pseudo labels for all samples, then aligned the intra-class distributions of two domains with the help of the predicted labels. Under the deep learning framework, the proposed algorithm effectively avoided label information being washed away and greatly improved the prediction ability on the target domain. The experimental results show that the proposed algorithm outperforms the traditional algorithms on the benchmarks. [ABSTRACT FROM AUTHOR]

Copyright of Application Research of Computers / Jisuanji Yingyong Yanjiu is the property of Application Research of Computers Edition and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Guangdong University of Technology is the property of Journal of Guangdong University of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Application Research of Computers / Jisuanji Yingyong Yanjiu is the property of Application Research of Computers Edition and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Convolutional neural network's high computational complexity often obstructs its widespread adhibition in real-time and low-power applications. The existing software implementation solution cannot meet the demands of the CNN for computing performance and power consumption. The traditional FPGA-oriented CNN construction method has problems such as complicated process, long cycle and small optimization space. For these problems, according to the characteristics of CNN calculation pattern, this paper proposed a design and scheduling mechanism of convolutional neural network accelerator for cloud FPGA. By using for reference the design which based HLS technology, importing the cyclic cutting parameters and rearranging the convolution layer circularly, it constructed the network in a modular way, and extended parameters to further optimize the accelerator processing process. It summarized the scheduling scheme by analyzing the characteristics of system tasks and resources, and optimized its design from two aspects of control and data flow. In comparison with other existing works, the proposed design provided a solution with flexibility, low energy consumption, high energy efficiency and performance. The design also discussed the efficient universal scheduling scheme of the accelerator. Experimental results show that the accelerator can improve the computing speed and reduce the power consumption. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Guangdong University of Technology is the property of Journal of Guangdong University of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Difficulty of crowd scene modeling, low reliability of the detection and so on problems, are main challenges of abnormal crowd detection. To address these challenges, this paper proposed a localized recurrent neural net-work based approach. Firstly, this method extracted MHOF( multi-scale histogran of optical flow) and used ii as the dynamic features for each mesh of the crow scene. Then, it applied localized recurrent neural network to model and predict the dynamic property of the crowd scene. Finally , it conducted abnormal monitoring and localization based on reconstruction error of the front and rear frame. Experimental results on the UCSDdatasets verify the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR]

Copyright of Application Research of Computers / Jisuanji Yingyong Yanjiu is the property of Application Research of Computers Edition and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Guangdong University of Technology is the property of Journal of Guangdong University of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Image & Graphics is the property of Editorial Office of Journal of Image & Graphics and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)