Abstract Location data management plays a crucial role in facilitating data collection and supporting location-based services. However, the escalating volume of transportation big data has given rise to increased concerns regarding privacy and security issues in data management, potentially posing threats to the lives and property of users. At present, there are two possible attacks in data management, namely Reverse-clustering Inference Attack and Mobile-spatiotemporal Feature Inference Attack. Additionally, the dynamic allocation of privacy budgets emerges as an NP-hard problem. To protect data privacy and maintain utility in data management, a novel protection model for location privacy information in data management, Classified Regional Location Privacy-Protection Model based on Personalized Clustering with Differential Privacy (PCDP-CRLPPM), is proposed. Firstly, a twice-clustering algorithm combined with gridding is proposed, which divides continuous locations into different clusters based on the different privacy protection needs of different users. Subsequently, these clusters are categorized into different spatiotemporal feature regions. Then, a Sensitive-priority algorithm is proposed to allocate privacy budgets adaptively for each region. Finally, a Regional-fuzzy algorithm is presented to introduce Laplacian noise into the centroids of the regions, thereby safeguarding users’ location privacy. The experimental results demonstrate that, compared to other models, PCDP-CRLPPM exhibits superior resistance against two specific attack models and achieves high levels of data utility while preserving privacy effectively.

Modern power systems are undergoing a paradigm shift from traditional grids towards smart grids. It fundamentally changes traditional power systems into complex cyber-physical systems. On the other hand, new challenges arise in terms of grid resilience, because natural disasters can cause damages to both cyber and physical systems. In this article, we propose a stochastic sequential restoration scheme for cyber-physical power distribution systems considering resilience. The sequential restoration problem is formulated as an uncertain Markov decision process (UMDP) with hurricanes incorporated as natural disasters. Different wind velocities and directions are considered as hurricane scenarios, which are used to obtain the fragility of distribution lines. The fragility functions are further used for the derivation of uncertain state transition functions of the UMDP. The minimax regret optimization considering the sample weights of UMDP is presented. The robust sequential actions are determined, such that the loads can be restored in a timely manner. To improve computational efficiency, a minimax regret policy iteration algorithm is presented based on the regret Bellman equation. Case studies are conducted based on the IEEE 123-Node Test Feeder and historical data of Hurricane Bonnie to demonstrate the effectiveness of the proposed scheme.

Recently, the machine unlearning has emerged as a popular method for efficiently erasing the impact of personal data in machine learning (ML) models upon the data owner's removal request. However, few studies take into consideration the security concerns that may exist in the unlearning process. In this article, we propose the first unlearning attack dubbed unlearning attack for regression learning (UnAR) to deliberately influence the predictive behavior of the target sample against regression learning models. The central concept of UnAR revolves around misleading the regression model into erasing the information associated with the influential samples for the target sample. Observing that the influential samples for target data are generally located far away from the regression plane, we thus propose two novel methods, known as influential sample selection (ISS) and influential sample unlearning (ISU), to identify and subsequently eliminate the lineage of the influential samples. By doing so, we can substantially introduce bias into the prediction pertaining to the target sample, yielding the deliberate manipulation for the user adversely. We extensively evaluate UnAR on five public datasets, and the experimental results indicate our attacks can achieve prediction deviations over 35% by unlearning only 0.5% data as the influential samples.

Deep neural networks have been successfully applied in many computer vision tasks. However, these models suffer catastrophic forgetting when learning new knowledge incrementally. To overcome the stability-plasticity dilemma, class incremental learning (CIL) has been widely discussed recently. The state-of-the-art CIL methods mainly leverage additional exemplar sets, thus memory costly and may raise privacy issues. To that end, we propose an autonomous generative feature replay (AGFR) framework without using exemplar sets. It consists of three modules: the feature extractor module, the feature generator module, and the unified classification module. First, to stabilize features over tasks, robust feature extractors are learned in a self-supervised manner and thus generalize well to unseen data. Second, instead of using exemplar sets or producing raw images, we propose an autonomous generative feature replay scheme to constantly update unified classifier in CIL without saving any image data. This strategy avoids overwhelming memory usage or poor quality of the generated raw images. Experiments demonstrate that our method achieves state-of-the-art performance in terms of average classification accuracy.⋆

Traditional shape descriptors have been gradually replaced by convolutional neural networks due to their superior performance in feature extraction and classification. The state-of-the-art methods recognize object shapes via image reconstruction or pixel classification. However, these methods are biased toward texture information and overlook the essential shape descriptions, thus, they fail to generalize to unseen shapes. We are the first to propose a few-shot shape descriptor (FSSD) to recognize object shapes given only one or a few samples. We employ an embedding module for FSSD to extract transformation-invariant shape features. Secondly, we develop a dual attention mechanism to decompose and reconstruct the shape features via learnable shape primitives. In this way, any shape can be formed through a finite set basis, and the learned representation model is highly interpretable and extendable to unseen shapes. Thirdly, we propose a decoding module to include the supervision of shape masks and edges and align the original and reconstructed shape features, enforcing the learned features to be more shape-aware. Lastly, all the proposed modules are assembled into a few-shot shape recognition scheme. Experiments on five datasets show that our FSSD significantly improves the shape classification compared to the state-of-the-art under the few-shot setting.

Earthquakes, which consist of one intensive main shock and a series of aftershocks, can significantly damage power distribution systems (PDSs). In this article, a data-driven PDS resilience enhancement strategy is proposed against multishocks of earthquakes. In particular, the investment and prepositioning of mobile emergency generators (MEGs) is determined against multishocks of earthquakes. The reallocation of MEG and the repair scheduling are obtained considering aftershocks and postrestoration failures. A resistibility index (RI) is developed based on hierarchical hidden Markov model (HHMM) for stochastic resilience evaluation. The historical earthquake data are incorporated into the HHMM as observed information of multishocks of earthquakes. Based on the RI, the problems of prepositioning and reallocation of MEGs are formulated as mixed-integer programming problems. The problem of repair scheduling is formulated as an adaptive multiperiod two-stage stochastic programming problem, for which a revision period is introduced to allow the decisions to adapt to the uncertainties after the revision. To reduce the computational complexity, an iterative algorithm is presented based on linear programming relaxation. The strategy is verified via case studies on the IEEE 123-Node Test Feeder and historical earthquake data. It shows by considering RI, the resilience of restoration can be optimized against future shocks of earthquakes. Also, the overall consideration of MEG investment, prepositioning, reallocation, and repair scheduling against multishocks of earthquakes can achieve an improved restoration performance.

Automatic and accurate detection of anatomical landmarks is an essential operation in medical image analysis with a multitude of applications. Recent deep learning methods have improved results by directly encoding the appearance of the captured anatomy with the likelihood maps (i.e., heatmaps). However, most current solutions overlook another essence of heatmap regression, the objective metric for regressing target heatmaps and rely on hand-crafted heuristics to set the target precision, thus being usually cumbersome and task-specific. In this paper, we propose a novel learning-to-learn framework for landmark detection to optimize the neural network and the target precision simultaneously. The pivot of this work is to leverage the reinforcement learning (RL) framework to search objective metrics for regressing multiple heatmaps dynamically during the training process, thus avoiding setting problem-specific target precision. We also introduce an early-stop strategy for active termination of the RL agent's interaction that adapts the optimal precision for separate targets considering exploration-exploitation tradeoffs. This approach shows better stability in training and improved localization accuracy in inference. Extensive experimental results on two different applications of landmark localization: 1) our in-house prenatal ultrasound (US) dataset and 2) the publicly available dataset of cephalometric X-Ray landmark detection, demonstrate the effectiveness of our proposed method. Our proposed framework is general and shows the potential to improve the efficiency of anatomical landmark detection.

The occurrence of natural disasters imposes potential damage on the power systems. Many methods have been proposed to improve power system resilience by constructing load restoration when the blackout happens. However, disaster like earthquakes may have secondary impacts (i.e., aftershock of earthquake) which would undermine the post-contingency power supply and cause restoration failure. To address such challenge, this paper proposes a Mobile Energy Resource (MER) allocation strategy to restore critical loads by determining the most reliable restoration path against the potential earthquakes. Benefited from the mobility of MER, the restoration paths can always be adjusted according to the real damage of earthquakes as long as there are available paths and capacity. On the other hand, when curtailment is inevitable, critical loads would be optimally shed based on priority. The effectiveness of the proposed strategy is verified by case studies on the modified IEEE 123-node test feeder.