Federated learning has emerged as a widely adopted distributed machine learning paradigm. In the field of medical diagnosis, it has become a research hotspot, enabling multiple institutions to collaboratively leverage medical data for accurate analysis. However, the distributed nature of federated learning also introduces new challenges in data security and privacy protection, i.e., a curious server might collude with the client to infer private data of honest clients. In this paper, we implement our own FHE library and integrate it with several widely used federated learning methods, providing a unified framework. Our framework employs multiple-party Fully Homomorphic Encryption (FHE) to remove the requirements of a trusted third party or central key servers and address data security and privacy concerns in federated learning, ensuring that the data of honest participants is never exposed while maintaining the accuracy of the final analysis. We deploy our Privacy-preserving Federated Learning framework in the context of Melanoma Diagnosis, PFL-MD, and across multiple types of widely used benchmark datasets, our method achieves accuracy comparable to that of the original federated learning framework, thereby enabling reliable melanoma diagnosis. Extensive experiments further demonstrate the effectiveness of our framework.

Deep learning models, due to their strong capabilities in learning from data and representing features, have been widely deployed, particularly in the medical field. However, given the limitations of medical devices and deployment scenarios, more and more researchers are focusing on how to better deploy deep learning models on resource-constrained edge devices to enable real-time medical data analysis. Nevertheless, such resourceconstrained IoT devices present significant challenges, making it difficult to achieve both high accuracy and low inference latency. To address this problem, we propose a novel framework, RC-DLM, designed to split and execute complex deep learning models on IoT devices. Specifically, we partition a deep learning model into multiple sub-models according to the computational capacity of each device, with each sub-model responsible for handling a subset of classes. To further reduce computation overhead and inference latency, we integrate a class-wise pruning method to shrink the size of each sub-model. Through large-scale experiments conducted on four popular datasets with three model architectures, we demonstrate that our approach significantly reduces inference latency and model size by up to 5.72 times and 57.5 times, respectively. We further deploy our method on real-world edge devices and compare it with state-of-theart approaches, evaluating the three most important aspects: accuracy, inference time, and model size. The comprehensive experimental results of our RC-DLM confirm the effectiveness of our proposed method.

In regions with limited medical resources, detecting melanoma has long been a challenging problem, particularly due to its heterogeneity and rarity. However, with the recent advances in deep learning technologies, it has become increasingly feasible to enable melanoma detection in remote areas. By collecting melanoma datasets, researchers have focused on exploring more effective models to learn discriminative features and deploy automated detection systems, with the aim of assisting doctors in remote or under-resourced regions. In this paper, we explore the latest deep learning model, Vision Mamba, as an AI-based diagnostic approach for melanoma detection. Furthermore, we enhance the Vision Mamba model by analyzing the differences between benchmark datasets and real-world clinical data, aiming to further reduce the false positive rate and minimize misdiagnosis. We conduct experiments on multiple types of widely used datasets across three model architectures and compare our approach against several popular baseline methods. Extensive experimental results demonstrate that Vision Mamba achieves state-of-the-art performance and validates its effectiveness for melanoma diagnosis in diverse clinical scenarios.

Deep learning models are increasingly utilized on resource-constrained edge devices for real-time data analytics. Recently, Vision Transformer and their variants have shown exceptional performance in various computer vision tasks. However, their substantial computational requirements and low inference latency create significant challenges for deploying such models on resource-constrained edge devices. To address this issue, we propose a novel framework, ED-ViT, which is designed to efficiently split and execute complex Vision Transformers across multiple edge devices. Our approach involves partitioning Vision Transformer models into several sub-models, while each dedicated to handling a specific subset of data classes. To further reduce computational overhead and inference latency, we introduce a class-wise pruning technique that decreases the size of each sub-model. Through extensive experiments conducted on five datasets using three model architectures and actual implementation on edge devices, we demonstrate that our method significantly cuts down inference latency on edge devices and achieves a reduction in model size by up to 28.9 times and 34.1 times, respectively, while maintaining test accuracy comparable to the original Vision Transformer. Additionally, we compare ED-ViT with two state-of-the-art methods that deploy CNN and SNN models on edge devices, evaluating metrics such as accuracy, inference time, and overall model size. Our comprehensive evaluation underscores the effectiveness of the proposed ED-ViT framework.

One-shot Federated Learning (OFL) is a distributed machine learning paradigm that constrains client-server communication to a single round, addressing privacy and communication overhead issues associated with multiple rounds of data exchange in traditional Federated Learning (FL). OFL demonstrates the practical potential for integration with future approaches that require collaborative training models, such as large language models (LLMs). However, current OFL methods face two major challenges: data heterogeneity and model heterogeneity, which result in subpar performance compared to conventional FL methods. Worse still, despite numerous studies addressing these limitations, a comprehensive summary is still lacking. To address these gaps, this paper presents a systematic analysis of the challenges faced by OFL and thoroughly reviews the current methods. We also offer an innovative categorization method and analyze the trade-offs of various techniques. Additionally, we discuss the most promising future directions and the technologies that should be integrated into the OFL field. This work aims to provide guidance and insights for future research.

Variable selection for genome-wide association studies is a key focus for bioinformatics researchers in high-performance computing. The rapid growth of biological and biomedical data demands has led to high-dimensional, heterogeneous datasets characterized by non-i.i.d. properties and numerous response variables, often resulting in false negatives or positives in recovered results. Traditional methods, when nal̈ively applied, yield suboptimal performance due to confounding factors. To account for the complex interdependencies in heterogeneous data and enhance the practical outcomes of genome-wide association studies, we introduce two methods, TGsLMM and TTsLMM, which balance effects between response and explanatory variables for subpopulation inference. Our unified framework performs sparse variable selection using graph-structured or tree-guided structures in a low-rank linear mixed model. Additionally, we extend our approach to high-dimensional datasets and adaptively select the covariance structure for genomic data. Extensive experiments on synthetic and three real-world datasets emphasize the robustness and effectiveness of our proposed methods, achieving the highest ROC area compared to baselines and superior results for future potential.

Multi-agent systems have been recognized as a promising solution for developing collaborative design and manufacturing systems. Resource sharing is one of the key issues that need to be addressed effectively in these systems. Effective resources utilization usually requires solving the synchronization interdependency problem among the agents. This paper addresses the synchronization interdependency problem and provides an effective solution for resource sharing at runtime. Also, it describes a proof-of-concept prototype implementation of the synchronization device using coordinated, intelligent rational agent model and the JADE platform.

The research objective is to develop methodologies and framework for collaborative process planning and scheduling, supported by a real-time monitoring system in distributed environments. A function block enabled collaborative process planning approach is proposed to handle various dynamic changes during process plan generation and execution. This paper focuses on collaborative process planning, particularly on the development of a function block designer As function blocks can sense environmental changes, it is expected that a so generated process plan can adapt itself to the changes with dynamically optimized solutions for plan execution and process monitoring.