It has been widely proven that Augmented Reality (AR) brings numerous benefits in learning experiences, including enhancing learning outcomes and motivation. However, not many studies investigate how different forms o...
详细信息
Graph neural networks (GNNs) have gained increasing popularity, while usually suffering from unaffordable computations for real-world large-scale applications. Hence, pruning GNNs is of great need but largely unexplor...
详细信息
Graph neural networks (GNNs) have gained increasing popularity, while usually suffering from unaffordable computations for real-world large-scale applications. Hence, pruning GNNs is of great need but largely unexplored. The recent work Unified GNN Sparsification (UGS) studies lottery ticket learning for GNNs, aiming to find a subset of model parameters and graph structures that can best maintain the GNN performance. However, it is tailed for the transductive setting, failing to generalize to unseen graphs, which are common in inductive tasks like graph classification. In this work, we propose a simple and effective learning paradigm, Inductive Co-Pruning of GNNs (ICPG), to endow graph lottery tickets with inductive pruning capacity. To prune the input graphs, we design a predictive model to generate importance scores for each edge based on the input. To prune the model parameters, it views the weight’s magnitude as their importance scores. Then we design an iterative co-pruning strategy to trim the graph edges and GNN weights based on their importance scores. Although it might be strikingly simple, ICPG surpasses the existing pruning method and can be universally applicable in both inductive and transductive learning settings. On 10 graph-classification and two node-classification benchmarks, ICPG achieves the same performance level with 14.26%–43.12% sparsity for graphs and 48.80%–91.41% sparsity for the GNN model.
Cloud computing solutions are becoming more and more popular as a way for organizations to improve productivity, save costs, and simplify procedures. The advantage of cloud services is that they enable users to store ...
详细信息
Industrial Internet of Things(IIoT)is a pervasive network of interlinked smart devices that provide a variety of intelligent computing services in industrial *** IIoT nodes operate confidential data(such as medical,tr...
详细信息
Industrial Internet of Things(IIoT)is a pervasive network of interlinked smart devices that provide a variety of intelligent computing services in industrial *** IIoT nodes operate confidential data(such as medical,transportation,military,etc.)which are reachable targets for hostile intruders due to their openness and varied *** Detection Systems(IDS)based on Machine Learning(ML)and Deep Learning(DL)techniques have got significant ***,existing ML and DL-based IDS still face a number of obstacles that must be *** instance,the existing DL approaches necessitate a substantial quantity of data for effective performance,which is not feasible to run on low-power and low-memory *** and fewer data potentially lead to low performance on existing *** paper proposes a self-attention convolutional neural network(SACNN)architecture for the detection of malicious activity in IIoT networks and an appropriate feature extraction method to extract the most significant *** proposed architecture has a self-attention layer to calculate the input attention and convolutional neural network(CNN)layers to process the assigned attention features for *** performance evaluation of the proposed SACNN architecture has been done with the Edge-IIoTset and X-IIoTID *** datasets encompassed the behaviours of contemporary IIoT communication protocols,the operations of state-of-the-art devices,various attack types,and diverse attack scenarios.
automating decision-making on a wide range of aspects are affecting humanity. There is great promise in machine learning (ML) and artificial intelligence (AI) algorithms for assisting with predicted and automated deci...
详细信息
ISBN:
(纸本)9798350330137
automating decision-making on a wide range of aspects are affecting humanity. There is great promise in machine learning (ML) and artificial intelligence (AI) algorithms for assisting with predicted and automated decision-making and transforming both societies and industries to improve African citizens' everyday lives. In a business context, platforms such as cloud computing can benefit a lot from AI and ML. Prediction models could be developed to predict the traffic flow of cloud computing resources, which could ease traffic flow. Although AI and ML cloud computing are powerful and vital, they are unfortunately not immune from ethical challenges. As well as improving the allocation of cloud computing resources with prediction models. In recent years, despite AI's powerful performance, in order to conduct responsible AI and ML modeling, it is important not to ignore ethical issues that might need serious investigation, particularly if human and classified data are concerned. In this regard, the objective of this paper is to determine the prevalence of AI and ML ethics research in Africa. In this paper, RStudio and the computing-related databases (IEEE Xplore, Web of service and Scopus) were used to conduct a bibliometric analysis on the prevalence of AI and ML ethics that relate to the cloud computing environment. The scope was minimized due to its prevalence in the African context. The findings show that AI and ML ethics are understudied on the African continent compared to the West. In this regard, there is a need to conduct more AI and ML ethics-related research in Africa because failure to do so will result in the West shaping the ethical discussion in areas of importance for African development, such as cloud computing. The latter will allow the West to slowly impose its perspective without considering African perspectives. Of note, this paper contributes to the doctoral study with an objective phrased as "to conduct bibliometric analysis to identify the preval
Association in-between features has been demonstrated to improve the representation ability of data. However, the original association data reconstruction method may face two issues: the dimension of reconstructed dat...
详细信息
Association in-between features has been demonstrated to improve the representation ability of data. However, the original association data reconstruction method may face two issues: the dimension of reconstructed data is undoubtedly higher than that of original data, and adopted association measure method does not well balance effectiveness and efficiency. To address above two issues, this paper proposes a novel association-based representation improvement method, named as AssoRep. AssoRep first obtains the association between features via distance correlation method that has some advantages than Pearson’s correlation coefficient. Then an improved matrix is formed via stacking the association value of any two features. Next, an improved feature representation is obtained by aggregating the original feature with the enhancement matrix. Finally, the improved feature representation is mapped to a low-dimensional space via principal component analysis. The effectiveness of AssoRep is validated on 120 datasets and the fruits further prefect our previous work on the association data reconstruction.
With the advent of data 3.0 and analytics 3.0, system thinkers are in the position to provide a bigger picture in datascience and data engineering. In the data life cycle, a system thinking approach emphasises data-d...
详细信息
Online streaming feature selection(OSFS),as an online learning manner to handle streaming features,is critical in addressing high-dimensional *** real big data-related applications,the patterns and distributions of st...
详细信息
Online streaming feature selection(OSFS),as an online learning manner to handle streaming features,is critical in addressing high-dimensional *** real big data-related applications,the patterns and distributions of streaming features constantly change over time due to dynamic data generation ***,existing OSFS methods rely on presented and fixed hyperparameters,which undoubtedly lead to poor selection performance when encountering dynamic *** make up for the existing shortcomings,the authors propose a novel OSFS algorithm based on vague set,named *** main idea is to combine uncertainty and three-way decision theories to improve feature selection from the traditional dichotomous method to the trichotomous ***-Vague also improves the calculation method of correlation between features and ***,OSFS-Vague uses the distance correlation coefficient to classify streaming features into relevant features,weakly redundant features,and redundant ***,the relevant features and weakly redundant features are filtered for an optimal feature *** evaluate the proposed OSFS-Vague,extensive empirical experiments have been conducted on 11 *** results demonstrate that OSFS-Vague outperforms six state-of-the-art OSFS algorithms in terms of selection accuracy and computational efficiency.
Reliable artificial intelligence (AI) systems not only propose a challenge on providing intelligent services with high quality for customers but also require customers' privacy to be protected as much as possible ...
详细信息
Graph neural networks(GNNs)have gained traction and have been applied to various graph-based data analysis tasks due to their high ***,a major concern is their robustness,particularly when faced with graph data that h...
详细信息
Graph neural networks(GNNs)have gained traction and have been applied to various graph-based data analysis tasks due to their high ***,a major concern is their robustness,particularly when faced with graph data that has been deliberately or accidentally polluted with *** presents a challenge in learning robust GNNs under noisy *** address this issue,we propose a novel framework called Soft-GNN,which mitigates the influence of label noise by adapting the data utilized in *** approach employs a dynamic data utilization strategy that estimates adaptive weights based on prediction deviation,local deviation,and global *** better utilizing significant training samples and reducing the impact of label noise through dynamic data selection,GNNs are trained to be more *** evaluate the performance,robustness,generality,and complexity of our model on five real-world datasets,and our experimental results demonstrate the superiority of our approach over existing methods.
暂无评论