In recent years, the attention of online cross-university courses has been increasing, and students in universities want to increase their knowledge and professional skills by taking online courses from different universities, which raises the issue of course credit verification. In the past, the credits obtained by students in online courses lack endorsement from the education department, and the students’ learning process could not be verified. Therefore, the credits of online courses in one university could not be recognized by other universities. The education departments of some countries and regions implement credit conversion rules to convert the credits obtained by students in online courses into university credits or certificates endorsed by the education department. However, these schemes rely too much on the authority of the education department, and the process of students obtaining credits cannot be verified. In addition, the centralized storage method makes the data of education departments at risk of leakage or tampering. With the emergence of blockchain technology, some researchers have proposed the use of blockchain to store students’ credits, making it possible to reach consensus among multiple parties on the blockchain while ensuring that credits are not tampered with, but these schemes cannot test the learning process of students and the recognition of credits still relies on the authority of the education department. To solve the above problems, this paper proposes a cross-university course learning system with verifiable credits based on Hyperledger Fabric consortium blockchain technology, and the consortium includes many universities. The credits obtained by students in the course and the hash value of the learning records are stored on the blockchain, and the data on the blockchain is jointly maintained by the universities in the system. One university can verify the homework and final examination of students to check the real ability of students, thus recognizing the credits from other universities, and at the same time, to protect the privacy of students, the important data of students are encrypted for transmission.

To improve the robustness of SVM models to noise and outliers, fuzzy support vector machine (FSVM) has been proposed. However, many existing FSVM models have limitations such as their dependence on assumptions, limited optimization, and unreasonable handling of noise. To address these problems, we propose a novel approach called curriculum learning-based FSVM. Our approach employs a curriculum-learning strategy, where model initially learns easy samples to avoid noise interference and obtain a good initial solution, before proceeding to learn all samples, including hard ones. To distinguish between easy and hard samples, we introduce an adaptive density-based clustering model and it is extended to kernel feature space. Moreover, we propose a slack variable-based fuzzy membership function to evaluate the importance of samples. Additionally, our model adaptively adapts the importance of samples based on feedback during the learning process. Finally, our experimental results on popular benchmarks demonstrate that our proposed model outperforms existing competitors in terms of accuracy and robustness.

In graph-structured data, the node content contains rich information. Therefore, how to effectively utilize the content is crucial to improve the performance of graph convolutional networks (GCNs) on various analytical tasks. However, current GCNs do not fully utilize the content, especially multi-order content. For example, graph attention networks (GATs) only focus on low-order content, while high-order content is completely ignored. To address this issue, we propose a novel graph attention network with adaptability that could fully utilize the features of multi-order content. Its core idea has the following novelties: First, we constructed a high-order content attention mechanism that could focus on high-order content to evaluate attention weights. Second, we propose a multi-order content attention mechanism that can fully utilize multi-order content, i.e., it combines the attention mechanisms of high- and low-order content. Furthermore, the mechanism has adaptability, i.e., it can perform a good trade-off between high- and low-order content according to the task requirements. Lastly, we applied this mechanism to constructing a graph attention network with structural symmetry. This mechanism could more reasonably evaluate the attention weights between nodes, thereby improving the convergence of the network. In addition, we conducted experiments on multiple datasets and compared the proposed model with state-of-the-art models in multiple dimensions. The results validate the feasibility and effectiveness of the proposed model.

We consider the problem of estimating lead times of jobs that arrive dynamically in a production shop. The lead time of a job refers to the duration between the job's arrival and completion. To improve accuracy of estimation and generality of application, we propose a new rule that takes into account parallelization of waiting queues in the system and job-specific sequence of visiting workstations. We carry out extensive experiments to evaluate performance of the rule, and the results suggest that the rule is high-performing in production shops including single-machine shops, parallel-machine shops, flexible flow shops, and flexible job shops. We further provide the mean of best values found for the parameters in the proposed rule under three performance measures: relative error ratio, mean tardiness ratio, and percentage of tardy jobs, respectively, to help users determine the parameters according to their needs.