Recent researches on data-driven and low-sparsity data injection attacks have been presented, respectively. To combine the two main goals (data-driven and low-sparsity) of research, this paper presents a data-driven and low-sparsity false data injection attack strategy. The proposed attacking strategy (EID: Eliminate-Infer-Determine) is divided into three stages. In the first step, the intercepted data is preprocessed by sparse optimization techniques to eliminate the outliers. The recovered data is then exploited to learn about the system matrix based on the parallel factorization algorithm in the second step. In the third step, the approximated system matrix is applied for the design of sparse attack vector based on the convex optimization. The simulation results show that the EID attack strategy achieves a better performance than the improved ICA-based attack strategy in constructing perfect sparse attack vectors. What is more, data-driven implementation of the proposed strategy is also presented which ensures attack performance even without the prior information of the system.

Background: The primary objective of this study is to ascertain whether the Expectation Confirmation Model can be expanded by external variables including computer anxiety, social interaction, and self-efficacy to better understand the intention to continue using online learning systems in the post-pandemic era among vocational college students. Moreover, this research argues that the intention to continue using online learning systems among students may be gender-sensitive.Methods: The researchers surveyed 482 students from eight vocational colleges in Jiangxi Province using a structured questionnaire. Partial Least Squares Structural equation modeling is used to verify the research model.Results: The outcomes demonstrate that the proposed model adequately explains the continuous use intention for online learning systems at a 76.6% confidence level. All of the newly introduced variables in the ECM are shown to be significant and relevant to explicate continuous use intention. Our survey results show that gender differences in intention to continue using online learning systems exist objectively, but this difference is not a natural difference.Conclusion: This research fills a void in the current literature on online learning and probes into how learning may be made more long-lasting in intricate environments.

The blind zone that accompanies the right-turn process of semitrailer trucks is a major cause of crashes and the high fatality of vulnerable road users (VRUs). Understanding the relationship between the blind zone and right-turn collisions will play a positive role in preventing such accidents. The purpose of this study was to investigate the formation of right-turn blind zones for semitrailer trucks and to determine the factors (turning speed, turning radius, and collision position) influencing the severity of accidents through real-world vehicle tests and PC-CRASH simulation. The results show that the calculation model of the inner wheel difference blind zone established for semitrailer trucks can provide more accurate estimation than the model for rigid trucks, due to the consideration of a virtual third axle between the tractor and the trailer. On the other hand, the PC-CRASH simulation test indicates the turning speed and turning radius directly affect the scale of the inner wheel difference blind zone, and larger blind zone and encroachment on adjacent lanes increase the potential for collision. Moreover, the difference in collision position is closely related to whether the rider suffers a secondary crush. Front position is more likely to cause the cyclist to be crushed. For further analysis, the long-term interaction between the blind zones resulting from the right rearview mirror and the inner wheel difference also increases the risk during a right turn. Therefore, reducing the blind zone in the right-turn process is the key to improving right-turn safety for semitrailer trucks and VRUs.

People’s attitudinal shifts toward an epidemic at different stages of the epidemic affect their travel behavior. Non-commuting travel behavior is more variable than commuting, as non-commuters have more travel options. However, few studies explored the changes in non-commuting travel and its influencing factors across different stages of sudden and localized COVID-19 outbreaks. Using survey data collected in Nanjing, China, where there was a sudden and localized outbreak of COVID-19, this research adopted the random parameter ordered logit model with heterogeneity in means and variances (HMV) to explore the factors influencing non-commuting travel in the early, middle, and late epidemic stages. The model results revealed that considering the HMV would improve the model fitness. In addition, the temporal stability of factors was investigated via a likelihood ratio test, which confirms traveler behavioral differences across different stages. The results showed that “e-bike ownership” and “the number of PCR (polymerase chain reaction) tests” is positively correlated with the number of non-commuting trips over three epidemic stages. The variables “people who live together have a red health code,”“mask replacement frequency,” and “risk-free areas” are significant in the early-stage and middle-stage models. The variables “people who live together have a green health code all the time” only become significant in the late-stage model. Research findings contribute to the understanding of non-commuting behaviors and targeted management needs during local outbreaks, and can help the government address travel issues under future major health events.