Federated learning (FL) typically faces data heterogeneity, i.e., distribution shifting among clients. Sharing clients' information has shown great potentiality in mitigating yet incurs a dilemma preserving privacy and promoting model performance. To alleviate the dilemma, we raise fundamental question: \textit{Is it possible to share partial features tackle heterogeneity?} In this work, give an affirmative answer question by proposing novel approach called {\textbf{Fed}erated...

Density-based clustering for big data is critical many modern applications ranging from Internet processing to massive-scale moving object management. This paper proposes Cludoop algorithm, an efficient distributed density-based using Hadoop. First, we propose a serial algorithm CluC by leveraging cell partition optimization and c-cluster fast find clusters. completes classification of the points relationships connected cells around instead expensive completed neighbor query, which...

The Internet of Things (IoT) boom has enabled Service Providers (ISPs) to collect an enormous amount high-dimensional data. Performing range queries on such data can effectively reuse them help ISPs offer better services. Owing the low cost and high resource utilization cloud computing, increasing number are inclined outsource services it. However, as is not fully trusted, need be encrypted before being outsourced, which inevitably hinders many query services, e.g., queries. Various schemes...

Glycation is a non-enzymatic process occurring inside or outside the host body by attaching sugar molecule to protein lipid molecule. It an important form of post-translational modification (PTM), which impairs function and changes characteristics proteins so that identification glycation sites may provide some useful guidelines understand various biological functions proteins. In this study, we proposed accurate prediction tool, named Glypre, for lysine glycation. Firstly, used multiple...

Abstract Currently, a novel of meta-heuristic algorithm called monarch butterfly optimization (MBO) is presented for solving machine learning and continuous problems. It has been proved experimentally that MBO superior to artificial bee colony (ABC), ant (ACO), Biogeography-based (BBO), differential evolution (DE) simple genetic (SGA) algorithms on most test functions. This paper presents new version with simulated annealing (SA) strategy SAMBO. The SA put in the migration operator adjusting...

Federated Learning (FL) models often experience client drift caused by heterogeneous data, where the distribution of data differs across clients. To address this issue, advanced research primarily focuses on manipulating existing gradients to achieve more consistent models. In paper, we present an alternative perspective and aim mitigate it generating improved local First, analyze generalization contribution training conclude that is bounded conditional Wasserstein distance between different...

As a key task in autonomous driving, 3D object detection based on LiDAR-camera fusion is expected to achieve more robust results by the complementarity of two sensors. However, non-trivial. An existing problem for this type detector that scale and receptive field LiDAR point features image are not matched, leading information deficiency or redundancy fusion. This paper proposes Point-based Pyramid Attention Fusion (PPAF) module solve problem. The PPAF learns corresponding points with matched...

The existing commodity Wi-Fi-based human gait recognition systems mainly focus on a single subject due to the challenges of multisubject walking monitoring. To tackle problem, we propose Wi-Diag, first abnormal diagnosis system that leverages only one pair off-the-shelf commercial Wi-Fi transceivers separate each subject's information and maintains an excellent performance when scenario changes. It is intelligent can release experienced doctor from heavy load work. Multisubject modeled as...

Federated semi-supervised learning (FSSL) has emerged as a powerful paradigm for collaboratively training machine models using distributed data with label deficiency. Advanced FSSL methods predominantly focus on single model each client. However, this approach could lead to discrepancy between the objective functions of labeled and unlabeled data, resulting in gradient conflicts. To alleviate conflict, we propose novel twin-model paradigm, called Twin-sight, designed enhance mutual guidance...

In this paper we propose an algorithm of computing minimal diagnosis based on BDD (Binary Decision Diagram). First give the concept disjunction equations, and map collection conflict sets into then compile system a BDD, further compute hitting by solving minimizing sets. The reliability completeness are proved. Variable ordering construct controlling used to reduce size final number non-minimal be generated. efficiency is highly improved. Experimental results show that isn't restricted...

As the development and application of Large Language Models (LLMs) continue to advance rapidly, enhancing their trustworthiness aligning them with human preferences has become a critical area research. Traditional methods rely heavily on extensive data for Reinforcement Learning from Human Feedback (RLHF), but representation engineering offers new, training-free approach. This technique leverages semantic features control LLM's intermediate hidden states, enabling model meet specific...

According to the characteristics of optimal elimination ordering problem in Bayesian networks, a heuristic-based genetic algorithm, cooperative coevolutionary framework and five grouping schemes are proposed. Based on these works, six algorithms constructed. Numerical experiments show that more robust than other existing swarm intelligence methods when solving problem.

With the rapid development of medical informatization as well larger quantities information and higher integration level, it has become a severe challenge to keep core data confidential. In this thesis, role next-generation firewall in network been first introduced. Then, designed an overall building scheme aimed at planned based on actual demand hospital. VLAN divided IP address each floor. Moreover, relevant protocols important equipment be used have introduced detail, including DHCP...

Abstract With the continuous progress of modern society, human rights are becoming more and perfect, especially medical system, because inevitability development, at this stage has achieved a lot proud results, but can not meet people’s pursuit health. Because use resources in industrial society led to emergence many viruses that have been seen many, so now kinds drugs, we considering things, one which is about adverse reactions drugs. Therefore, paper based on data mining algorithm monitor...

To solve the problem of searching for an optimal elimination ordering Bayesian networks, a novel effective heuristic, MinSum Weight, and ACS approach incorporated with multi-heuristic mechanism are proposed. The named MHC-ACS utilizes set heuristics to direct ants moving in search space. cooperation multiple helps explore more regions. Moreover, most appropriate heuristic will be identified reinforced evolution whole system. Experiments demonstrate that has better performance than other...

Nowadays, how to design a city with more sustainable features has become center problem in the field of social development, meanwhile it provided broad stage for application artificial intelligence theories and methods. Because is essentially constraint optimization problem, swarm algorithm extensive research natural candidate solving problem. TLBO (Teaching-Learning-Based Optimization) new algorithm. Its inspiration comes from "teaching" "learning" behavior teaching class life. The...