Structural reliability analysis is essential for evaluating system failure probabilities under uncertainties, yet it often faces computational efficiency challenges. While surrogate model-based techniques, including Kriging, are known for their high accuracy and efficiency, they typically employ a sequential learning strategy, which limits their potential for parallel computation. This article introduces the Local Penalization Adaptive Learning (LP-AL) method, which facilitates parallel adaptive reliability analysis; LP-AL introduces a penalty function that emulates the process of sequential learning strategies, thereby achieving parallelization. The method also integrates a global error-based stopping criterion and a sample pool reduction strategy to enhance efficiency. We tested LP-AL with five commonly used learning functions across various engineering scenarios. The results demonstrate that LP-AL achieves high accuracy and significantly reduces computational costs, making it a viable approach for diverse structural reliability analysis tasks.

Mobile edge computing (MEC) enables mobile devices (MDs) to offload their computational tasks to the network edge, significantly reducing transmission delay and energy consumption. In this paper, we develop a novel partial NOMA (PNOMA) based task offloading scheme in a multi-layer mobile computing network (MD-MEC-Cloud). PNOMA combines the high throughput of NOMA with the low interference of OMA for efficient, low-latency transmission. Furthermore, the PNOMA-based multi-layer collaborations enable rapid task processing across various computing requirements. We formulate a non-convex mixed-integer optimization problem aimed at minimizing the average delay across all MDs. To address this challenging problem, we propose an algorithm called reincarnating proximal policy optimization (RPPO), which uses online inference solutions to significantly reduce complexity. In addition, we incorporate accumulated apriori information into RPPO for fast retraining and design both a reward function and an evaluation phase to ensure the communication/computation constraints are met with a high probability. Simulation results demonstrate that the proposed task offloading scheme outperforms existing methods.

3D medical image segmentation is an essential task in the medical image field, which aims to segment organs or tumours into different labels. A number of issues exist with the current 3D medical image segmentation task: existing models cannot simultaneously obtain the space correlation and depth correlation of 3D slices; previous models suffer from local detail loss of positional embedding in 3D images; previous approaches often have blurring of boundaries in segmenting 3D images. To solve these shortcomings, we propose a 3D medical image segmentation model named TPM-SAM. In our model, we design a twinchannel image encoder to simultaneously capture the space correlation and depth correlation of 3D slices through a multi-head attention mechanism and improved adapters. Furthermore, we design a prompt encoding generator, which divides the volumetric image into small blocks and better captures the local detail information. In addition, we introduce a multi-layer aggregation decoder by employing U-Net with multi-level skip connection to solve the blurring of boundaries in processing 3D images. Finally, we experimented and evaluated our model on KiTS21 and LiTS17 datasets to compare with other baseline models.

The ever-increasing number of ubiquitous Internet of Things (IoT) applications entails a high demand for scarce communication and network resources. To meet this stringent requirement, mobile edge computing (MEC) is envisioned as a transformative technique to significantly streamline the existing network operations. Recently, device-to-device (D2D) communication has been proposed as a promising technology in 5G and beyond networks with a significantly increased transmission efficiency, especially suitable for small-packet task exchanges. In this paper, we incorporate D2D communication into the multi-layer computing network and propose a full-dimensional task offloading scheme by jointly optimizing task offloading decisions and computation/communication resource allocation. We formulate it as mixed-integer nonlinear programming (MINLP) problem, where the optimal branch-and-bound (B&B) algorithm with the full strong branching (FSB) variable selection policy features an extremely high complexity. To address this challenge, we propose inverse reinforcement learning with graph neural networks (GIRL) to generate a new variable selection policy that closely matches the FSB variable selection. Without sacrificing the global optimality, the GIRL can directly infer the variable selection with a much lower complexity, significantly accelerating the original B&B algorithm. Simulation results show that the GIRL achieves a lower complexity without sacrificing the global optimality. Furthermore, our proposed full-dimensional task offloading scheme achieves better performance than the existing schemes in terms of average delay for all mobile devices (MDs).

Cell-free massive multiple-input multiple-output (MIMO) deploys a large number of distributed access points (APs) without cell edges, offering seamless connectivity with significantly increased spectral efficiency and system capacity, but suffering degraded energy efficiency. In this paper, we develop a green energy scheme by simultaneously optimizing power allocation and AP selection. We formulate it as a non-convex mixed-integer nonlinear programming problem (MINLP), which is NP-hard. To address this challenging problem, we propose a learning-based algorithm that embeds non-convex optimization into contemporary deep reinforcement learning (DRL), referred to as optimization-embedded soft actor-critic with graph transformer networks (OSAC-G). OSAC-G enjoys the benefits of directly online inferring solutions for the non-convex problem with a much lower computational complexity compared to conventional non-convex optimization. Simulation results demonstrate that the green energy scheme significantly decreases energy consumption compared to the existing ones.

The rapid development of Internet of Things (IoT) applications requires efficient computing and communication resource allocation strategies to streamline the existing network operations. These strategies could be formulated as mixed-integer nonlinear programming (MINLP) problems, where the optimal branch-and-bound (B&B) with the full strong branching (FSB) variable selection policy features an extremely high complexity. We propose inverse reinforcement learning with graph neural networks (GNNIRL) to generate a new variable selection policy that closely matches the FSB variable selection. Without sacrificing the optimality, the GNNIRL can directly infer the variable selection with a significantly lower complexity, which is also verified by simulation.

The fine-grained image classification task is a major task in computer vision. Although many deep learning inter-pretable models have been proposed for this task, the accuracy and interpretability of these models need to be improved. We propose an interpretable fine-grained image classification model based on an improved neural prototype tree. In our model, we design the new multi-grained feature extraction network with three new backbone networks to extract features of fine-grained and multi-grained images more effectively. Furthermore, we design a new background prototype removing mechanism in the soft neural binary decision tree layer to optimize prototype path decision. Afterwards, we design a new loss function with both a leaf node loss function and a fully connected layer loss function to improve the generalization ability. Finally, we evaluate our model on three public datasets CUB-200-2011, FGVC-Aircraft, and Chest X-ray to compare with other baseline models.

Commonsense question answering (CSQA) is a challenging learning task that aims to give correct answers to commonsense questions. CSQA models combining large pretrained language models with knowledge graphs are proposed to perform one-way or two-way information fusion to enhance their commonsense reasoning ability. However, existing CSQA models only fuse local information at the word level, ignoring the global semantic information fusion. Furthermore, current CSQA models often introduce noise nodes when constructing the knowledge subgraph. In addition, existing methods neglect the edge information in message aggregation. To solve these shortcomings, we propose a novel CSQA model named MDEQA. In our model, we design the multi-layer attention fusion module to bidirectionally fuse the word-level local information and global semantic information of question context and knowledge subgraph. Moreover, we design the dynamic graph neural network module with improved GAT and aggregating edge information to form the dynamic subgraphs which alleviate the interference of noise nodes on reasoning and enhance the commonsense reasoning ability of our model. Finally, we evaluated our model on CommonsenseQA and OpenBookQA datasets to compare with other baseline models.

Medical image classification is a main task in medical diagnosis field. Some black box models have achieved expert-level accuracy on medical datasets, but these models are less adopted in clinical practice due to the lack of interpretability. In the past few years, designing interpretable models has been one of the major challenges in the medical field. The existing interpretable prototype networks only use medical images for training, ignoring the role of medical reports, and these medical reports can assist in prototype training and activation. Furthermore, existing prototype network methods neglect the position information in medical images, which is helpful for disease diagnosis. To solve these shortcomings, we propose an interpretable medical image classification framework (MProtoNet) that improves the accuracy and interpretability of disease predictions. In MProtoNet, we design a multimodal attention module and use prototype activation restriction loss to provide evidence for prototype training and activation. In addition, we design a position embedding module and multi-factor similarity calculation method to effectively utilize the position information in the image. We conducted experiments on the chest datasets MIMIC-CXR and open-I to test the model and compare it with other baseline models. Experimental results show that MProtoNet has made improvements in accuracy while preserving the interpretability of the model.

The significant success of deep learning has sparked interest in its application in medical diagnosis. Some deep learning models have achieved expert-level accuracy on some medical datasets, but these models are rarely used in clinical practice due to the lack of interpretability. Therefore, the research topic of explainable artificial intelligence (XAI) has emerged to make the reasoning process of the model transparent and interpretable. In this case, we applied interpretable artificial intelligence to dermatoscopy image diagnosis for the first time. Specifically, we use an interpretable prototype network for dermoscopy image diagnosis. To solve the problem of weak generalization performance of a single network, we propose to construct a new prototype network using the dual-path network. Besides, we propose a new gate similarity calculation method to reduce the activation of low-similarity regions, thereby reducing the generation of inaccurate prototypes and improving the diagnostic ability of the model. We conducted experiments on the dermoscopy datasets HAM10000 to test the model and compare it with other baseline models. Experimental results show that DP-ProtoNet has made improvements in accuracy while preserving the interpretability of the model.

Transcriptome prediction from genetic variation data is an important task in the privacy-preserving and biometrics field, which can better protect genomic data and achieve biometric recognition through transcriptome. Many transcriptome prediction methods have achieved good accuracy from genetic variation data. However, these traditional transcriptome prediction methods have the problems of linear assumption, overfitting, expose personal privacy, and extensive manual optimization. To solve these shortcomings, we propose an attention-based transcriptome prediction model from genetic variation named RATPM that improves the accuracy of transcriptome prediction and protects participant genomic data. In RATPM, we introduce and improve the deep learning model with multi-head self-attention into the transcriptome prediction stage of Predixcan, which uncovers the non-linear relationship between genetic variation and transcriptome. Moreover, we introduce a residual attention module to generate attention-aware features and extract more accurate features at different levels from genetic variation. Furthermore, we introduce the BERT pre-training module to encode genetic variation fully utilizing their contextual information. Our research enables scientific institutions to publish only predicted transcriptomic data for biometric purposes, thus protecting the genomic information of the subjects. Finally, we evaluated our model on the 1000 Genomes and Geuvadis projects datasets to compare with other baseline models.

In relation to the lack of design and innovative approach in conceptual design for the current prototype system, a parallel evolutionary strategy based on self-adaptive learning mechanism is developed to realize generative design. The 3-layer model named community-population-individual is proposed by this model. As the data structure for collaborations between subpopulations, the Blackboard model is introduced. And three learning operators are designed, through which combines the advantages of parallel evolutionary and genetic learning that improves the performance of traditional evolutionary strategy effectively. The application shows that this model can perfectly improve the rate and level of design innovation, stimulate creative inspiration for designers.