Hierarchical Optimization Method for Federated Learning with Feature Alignment and Decision Fusion

In the realm of data privacy protection,federated learning aims to collaboratively train a global model.However,heterogeneous data between clients presents challenges,often resulting in slow convergence and inadequate accuracy of the global model.Utilizing shared feature representations alongside customized classifiers for individual clients emerges as a promising personalized solution.Nonetheless,previous research has frequently neglected the integration of global knowledge into local representation learning and the synergy between global and local classifiers,thereby limiting model performance.To tackle these issues,this study proposes a hierarchical optimization method for federated learning with feature alignment and the fusion of classification decisions(FedFCD).FedFCD regularizes the relationship between global and local feature representations to achieve alignment and incorporates decision information from the global classifier,facilitating the late fusion of decision outputs from both global and local classifiers.Additionally,FedFCD employs a hierarchical optimization strategy to flexibly optimize model parameters.Through experiments on the Fashion-MNIST,CIFAR-10 and CIFAR-100 datasets,we demonstrate the effectiveness and superiority of FedFCD.For instance,on the CIFAR-100 dataset,FedFCD exhibited a significant improvement in average test accuracy by 6.83%compared to four outstanding personalized federated learning approaches.Furthermore,extended experiments confirm the robustness of FedFCD across various hyperparameter values.

Ulva prolifera subpixel mapping with multiple-feature decision fusion

The unavoidable nature of Ulva prolifera mixed pixel in low-resolution remote sensing images would result in rough boundary of U.prolifera patches,omission of tiny patches,and overestimation of coverage area.The decomposition of U.prolifera mixed pixel addresses the issue of coverage area overestimation,and the remaining problems can be alleviated by subpixel mapping(SPM).Due to the drift and dissipation of U.prolifera,a suitable SPM method is the single image-based unsupervised method.However,the method has difficulties in detail reconstruction,insufficient learning of spectral information,and SPM error introduced by abundance deviation.Therefore,we proposed a multiple-feature decision fusion SPM(MFDFSPM)method.It involves three branches to obtain the spatial,abundance,and spectral features of U.prolifera while considers multi-feature information using the fusion strategy.Experiments on the Geostationary Ocean Color Imager images in the Yellow Sea of China indicate that the MFDFSPM overperforms several typical U.prolifera SPM methods in higher accuracy and stronger robustness in both SPM and abundance calculation,which produced subpixel map with more detailed spatial information and less noise.

Synergy Decision for Radar and IRST Data Fusion

A new synergy decision method for radar and infrared search and track (IRST) data fusion is proposed, to solve such problems as how to decrease opportunities for radar suffering from being locked on by adverse electronic support measures (ESM), how to retrieve range information of the target during radar off, and how to detect the maneuver of the target. Firstly, polynomials used to predict target motion states are constructed. Secondly, a set of discriminants for detecting target maneuver are established by comparing the predicted values with the observations from IRST. Thirdly, a set of decisions are presented. Lastly, simulation is performed on the given scenario to test the validity of the method.

Optimal decision fusion given sensor rules

When all the rules of sensor decision are known, the optimal distributeddecision fusion, which relies only on the joint conditional probability densities, can be derivedfor very general decision systems. They include those systems with interdependent sensorobservations and any network structure. It is also valid for m-ary Bayesian decision problems andbinary problems under the Neyman-Pearson criterion. Local decision rules of a sensor withcommunication from other sensors that are optimal for the sensor itself are also presented, whichtake the form of a generalized likelihood ratio test. Numerical examples are given to reveal someinteresting phenomena that communication between sensors can improve performance of a senordecision, but cannot guarantee to improve the global fusion performance when sensor rules were givenbefore fusing.

Cloud-Based Diabetes Decision Support System Using Machine Learning Fusion

Diabetes mellitus,generally known as diabetes,is one of the most common diseases worldwide.It is a metabolic disease characterized by insulin deciency,or glucose(blood sugar)levels that exceed 200 mg/dL(11.1 ml/L)for prolonged periods,and may lead to death if left uncontrolled by medication or insulin injections.Diabetes is categorized into two main types—type 1 and type 2—both of which feature glucose levels above“normal,”dened as 140 mg/dL.Diabetes is triggered by malfunction of the pancreas,which releases insulin,a natural hormone responsible for controlling glucose levels in blood cells.Diagnosis and comprehensive analysis of this potentially fatal disease necessitate application of techniques with minimal rates of error.The primary purpose of this research study is to assess the potential role of machine learning in predicting a person’s risk of developing diabetes.Historically,research has supported the use of various machine algorithms,such as naïve Bayes,decision trees,and articial neural networks,for early diagnosis of diabetes.However,to achieve maximum accuracy and minimal error in diagnostic predictions,there remains an immense need for further research and innovation to improve the machine-learning tools and techniques available to healthcare professionals.Therefore,in this paper,we propose a novel cloud-based machine-learning fusion technique involving synthesis of three machine algorithms and use of fuzzy systems for collective generation of highly accurate nal decisions regarding early diagnosis of diabetes.

Multiscale Feature Fusion for Gesture Recognition Using Commodity Millimeter-Wave Radar

Gestures are one of the most natural and intuitive approach for human-computer interaction.Compared with traditional camera-based or wearable sensors-based solutions,gesture recognition using the millimeter wave radar has attracted growing attention for its characteristics of contact-free,privacy-preserving and less environmentdependence.Although there have been many recent studies on hand gesture recognition,the existing hand gesture recognition methods still have recognition accuracy and generalization ability shortcomings in shortrange applications.In this paper,we present a hand gesture recognition method named multiscale feature fusion(MSFF)to accurately identify micro hand gestures.In MSFF,not only the overall action recognition of the palm but also the subtle movements of the fingers are taken into account.Specifically,we adopt hand gesture multiangle Doppler-time and gesture trajectory range-angle map multi-feature fusion to comprehensively extract hand gesture features and fuse high-level deep neural networks to make it pay more attention to subtle finger movements.We evaluate the proposed method using data collected from 10 users and our proposed solution achieves an average recognition accuracy of 99.7%.Extensive experiments on a public mmWave gesture dataset demonstrate the superior effectiveness of the proposed system.

Scheme of Cooperative Spectrum Sensing Based on Adaptive Decision Fusion Algorithm

Spectrum sensing is one of the core technologies for cognitive radios （CR）, where reliable detection of the signals of primary users （PUs） is precondition for implementing the CR systems. A cooperative spectrum sensing scheme based on an adaptive decision fusion algorithm for spectrum sensing in CR is proposed in this paper. This scheme can estimate the PU prior probability and the miss detection and false alarm probabilities of various secondary users （SU）, and make the local decision with the Chair-Varshney rule so that the decisions fusion can be done for the global decision. Simulation results show that the false alarm and miss detection probabilities resulted from the proposed algorithm are significantly lower than those of the single SU, and the performance of the scheme outperforms that of the cooperative detection by using the conventional decision fusion algorithms.

Dependent Randomization in Parallel Binary Decision Fusion

We consider a parallel decentralized detection system employing a bank of local detectors(LDs)to access a commonly-observed phenomenon.The system makes a binary decision about the phenomenon,accepting one of two hypotheses(H_(0)("absent")or H_(1)("present")).The kth LD uses a local decision rule to compress its local observations yk into a binary local decision uk;uk=0 if the kth LD accepts H_(0)and uk=1 if it accepts H_(1).The kth LD sends its decision uk over a noiseless dedicated channel to a Data Fusion Center(DFC).The DFC combines the local decisions it receives from n LDs(u_(1),u_(2),...,u_(n))into a single binary global decision u_(0)(u_(0)=0 for accepting H_(0)or u_(0)=1 for accepting H_(1)).If each LD uses a single deterministic local decision rule(calculating uk from the local observations yk)and the DFC uses a single deterministic global decision rule(calculating u_(0)from the n local decisions),the team receiver operating characteristic(ROC)curve is in general non-concave.The system's performance under a Neyman-Pearson criterion may then be suboptimal in the sense that a mixed strategy may yield a higher probability of detection when the probability of false alarm is constrained not to exceed a certain value,α>0.Specifically,a"dependent randomization"detection scheme can be applied in certain circumstances to improve the system's performance by making the ROC curve concave.This scheme requires a coordinated and synchronized action between the DFC and the LDs.In this study,we specify when dependent randomization is needed,and discuss the proper response of the detection system if synchronization between the LDs and the DFC is temporarily lost.

DECISION FUSION FOR WIRELESS SENSOR NETWORKS UNDER NAKAGAMI FADING CHANNELS

Decision fusion rules for Wireless Sensor Networks (WSNs) under Nakagami fading channels are investigated in this paper. Considering the application limitation of Likelihood Ratio Test fusion rule based on information of Channel Statistics using Series expansion (LRT-CSS),and the detection performance limitation of the Censoring based Mixed Fusion rule (CMF),a new LRT fusion rule based on information of channel statistics has been presented using Laplace approximation (LRT-CSL). Theoretical analysis and simulations show that the proposed fusion rule provides better detection performance than the Censoring based Mixed Fusion (CMF) and LRT-CSS fusion rules. Furthermore,compared with LRT-CSS fusion rule,the proposed fusion rule expands the application range of likelihood ratio test fusion rule.

Multi-Model Fusion Framework Using Deep Learning for Visual-Textual Sentiment Classification

Multimodal Sentiment Analysis(SA)is gaining popularity due to its broad application potential.The existing studies have focused on the SA of single modalities,such as texts or photos,posing challenges in effectively handling social media data with multiple modalities.Moreover,most multimodal research has concentrated on merely combining the two modalities rather than exploring their complex correlations,leading to unsatisfactory sentiment classification results.Motivated by this,we propose a new visualtextual sentiment classification model named Multi-Model Fusion(MMF),which uses a mixed fusion framework for SA to effectively capture the essential information and the intrinsic relationship between the visual and textual content.The proposed model comprises three deep neural networks.Two different neural networks are proposed to extract the most emotionally relevant aspects of image and text data.Thus,more discriminative features are gathered for accurate sentiment classification.Then,a multichannel joint fusion modelwith a self-attention technique is proposed to exploit the intrinsic correlation between visual and textual characteristics and obtain emotionally rich information for joint sentiment classification.Finally,the results of the three classifiers are integrated using a decision fusion scheme to improve the robustness and generalizability of the proposed model.An interpretable visual-textual sentiment classification model is further developed using the Local Interpretable Model-agnostic Explanation model(LIME)to ensure the model’s explainability and resilience.The proposed MMF model has been tested on four real-world sentiment datasets,achieving(99.78%)accuracy on Binary_Getty(BG),(99.12%)on Binary_iStock(BIS),(95.70%)on Twitter,and(79.06%)on the Multi-View Sentiment Analysis(MVSA)dataset.These results demonstrate the superior performance of our MMF model compared to single-model approaches and current state-of-the-art techniques based on model evaluation criteria.

Machine Learning for Data Fusion:A Fuzzy AHP Approach for Open Issues

Data fusion generates fused data by combining multiple sources,resulting in information that is more consistent,accurate,and useful than any individual source and more reliable and consistent than the raw original data,which are often imperfect,inconsistent,complex,and uncertain.Traditional data fusion methods like probabilistic fusion,set-based fusion,and evidential belief reasoning fusion methods are computationally complex and require accurate classification and proper handling of raw data.Data fusion is the process of integrating multiple data sources.Data filtering means examining a dataset to exclude,rearrange,or apportion data according to the criteria.Different sensors generate a large amount of data,requiring the development of machine learning(ML)algorithms to overcome the challenges of traditional methods.The advancement in hardware acceleration and the abundance of data from various sensors have led to the development of machine learning(ML)algorithms,expected to address the limitations of traditional methods.However,many open issues still exist as machine learning algorithms are used for data fusion.From the literature,nine issues have been identified irrespective of any application.The decision-makers should pay attention to these issues as data fusion becomes more applicable and successful.A fuzzy analytical hierarchical process(FAHP)enables us to handle these issues.It helps to get the weights for each corresponding issue and rank issues based on these calculated weights.The most significant issue identified is the lack of deep learning models used for data fusion that improve accuracy and learning quality weighted 0.141.The least significant one is the cross-domain multimodal data fusion weighted 0.076 because the whole semantic knowledge for multimodal data cannot be captured.

Multisensor Information Fusion for Condition Based Environment Monitoring

Destructive wildfires are becoming an annual event,similar to climate change,resulting in catastrophes that wreak havoc on both humans and the envir-onment.The result,however,is disastrous,causing irreversible damage to the ecosystem.The location of the incident and the hotspot can sometimes have an impact on earlyfire detection systems.With the advancement of intelligent sen-sor-based control technologies,the multi-sensor data fusion technique integrates data from multiple sensor nodes.The primary objective to avoid wildfire is to identify the exact location of wildfire occurrence,allowingfire units to respond as soon as possible.Thus to predict the occurrence offire in forests,a fast and effective intelligent control system is proposed.The proposed algorithm with decision tree classification determines whetherfire detection parameters are in the acceptable range and further utilizes a fuzzy-based optimization to optimize the complex environment.The experimental results of the proposed model have a detection rate of 98.3.Thus,providing real-time monitoring of certain environ-mental variables for continuous situational awareness and instant responsiveness.

Fusion Strategy for Improving Medical Image Segmentation

In this paper,we combine decision fusion methods with four metaheuristic algorithms(Particle Swarm Optimization(PSO)algorithm,Cuckoo search algorithm,modification of Cuckoo Search(CS McCulloch)algorithm and Genetic algorithm)in order to improve the image segmentation.The proposed technique based on fusing the data from Particle Swarm Optimization(PSO),Cuckoo search,modification of Cuckoo Search(CS McCulloch)and Genetic algorithms are obtained for improving magnetic resonance images(MRIs)segmentation.Four algorithms are used to compute the accuracy of each method while the outputs are passed to fusion methods.In order to obtain parts of the points that determine similar membership values,we apply the different rules of incorporation for these groups.The proposed approach is applied to challenging applications:MRI images,gray matter/white matter of brain segmentations and original black/white images Behavior of the proposed algorithm is provided by applying to different medical images.It is shown that the proposed method gives accurate results;due to the decision fusion produces the greatest improvement in classification accuracy.

A NEW APPROACH TO THE DESIGN OF A SYSTEM FOR FEATURE EXTRACTION BASED ON HUMAN-COMPUTER COLLABORATIVE TACTIC

We are involved in an embarrassing situation that the limited capability of automated feature extraction in digital photogrammetric systems cannot satisfy the increasing needs for rapid acquisition of semantic information for applications. Facing this challenge, a new tactic, Human-Computer Collaborative (HCC) tactic, and a corresponding new method, Operator-Object Directed (OOD) method, are proposed for the design of a system for feature extraction from large scale aerial images. We hold that in almost all technical complex systems, full automation will be neither technically feasible nor socially acceptable. The system should be designed to optimize through the cooperative operation with two agents in the system: the hurtan and the computer.

基于自适应纹理特征融合的纹理图像分类方法

现有基于深度学习的图像分类方法普遍缺少纹理特征的针对性,分类精度较低,难以同时适用于简单纹理和复杂纹理分类。提出一种基于自适应纹理特征融合的深度学习模型,能够结合类间差异性纹理特征做出分类决策。首先,根据纹理特征的最大类间差异性,构建图像的纹理特征图像;然后,采用原始图像与特征鲜明的纹理特征图像并行训练改进的双线性模型,获取双通道特征;最后,基于决策融合构建自适应分类模块,连接原图与纹理集的平均池化特征图进行通道权重提取,根据通道权重融合2个并行神经网络模型的分类向量,得到最优融合分类结果。在KTH-TIPS,KTH-TIPS-2b, UIUC和DTD 4个公共纹理数据集上对模型的分类性能进行评估,分别得到了99.98%、99.95%、99.99%和67.09%的准确率,表明所提模型具有普遍高效的识别性能。

基于改进Grabcut分割与多特征决策融合的电力线放电痕迹识别

电力线触树故障中,导线表面的遗留痕迹是事故防治和责任认定的重要依据,但目前中外针对触树后电力线放电痕迹特征规律和辨识方法的研究极其匮乏。为此,搭建10 kV中压线路触树放电实验平台,采集放电后的导线表面痕迹图像,并对导线表面痕迹特征进行系统分析,为人工巡检和智能化痕迹识别提供基础依据。然后,提出改进型Grabcut前景提取方法,综合利用U^(2)Net的自动分割特点和Grabcut的高精度优势,解决Grabcut算法中初始框无法自动确定的问题,实现复杂背景下导线痕迹区域自动精准分割。最后,提出基于低层纹理、颜色特征和高层深度特征的导线表面痕迹全面表征,并采用多数投票规则实现低层和高层特征识别结果决策融合,获得导线痕迹辨识结果,测试实验中平均识别准确率达到91.68%,证明了方法的有效性。

激光雷达和相机的决策级融合目标检测方法

【目的】激光雷达与相机这两类传感器检测数据格式不统一、分辨率不同,且数据级和特征级的融合计算复杂度高,故提出一种决策级的目标融合检测方法。【方法】对激光雷达与相机的安装位置进行联合标定,实现这两类传感器检测结果的坐标系转换;利用匈牙利算法将激光雷达点云检测目标框和相机图像检测目标框进行匹配,设定目标框重合面积阈值,检测获得目标物的位置、类型等。【结果】实车测试结果表明,根据检测目标检测框长宽比选取不同交并比阈值的方法使得车辆和行人的目标识别准确率分别提升了3.3%和5.3%。利用公开数据集KITTI对所提融合方法进行验证,结果表明,在3种不同难度等级场景下,所提融合方法的检测精度分别达到了75.42%、69.71%、63.71%,与现有常用的融合方法相比,检测精度均有所提升。【结论】这两类传感器的检测目标框重合面积阈值对决策级融合检测结果影响较大,根据检测目标检测框长宽比选取不同阈值可有效提升车辆和行人的目标识别准确率。决策级融合方法能准确匹配雷达和相机的检测目标,有效提升目标检测精度。

人机决策融合复杂性机理及应用研究

在雷达组网探测群协同运用中,指战员一直难以快速正确地决策构群形态与管控预案,也就带来敏捷构群难与精准管控难问题,制约了空天新威胁的协同探测效能。文中在简述人机决策融合原理与复杂性的基础上,揭示了人机决策融合技术生成协同探测复杂性与破解空天新威胁突防复杂性的机理,建立了人机融合决策在协同探测全过程中的作用模型,设计了生成协同探测复杂性的应用流程,使雷达组网探测群具备“决策优化、构群敏捷、管控精准”等协同探测能力及“设变、知变、应变、求变”智能化博弈能力,能实现“基于任务敏捷构群、基于预案精准管控、基于情景快准微调”的协同探测。

基于多信息融合的换流站直流设备健康状态决策树评价模型

针对换流站直流设备间交互信息融合能力较差、小样本与大样本难以均衡的问题,设计基于多信息融合的换流站直流设备健康状态决策树评价模型。使用同质多传感器多信息数据融合方法对多信息实施融合处理;利用k近邻方法和层次聚类方法划分换流站直流设备健康状态量,提取健康状态量的有效信息作为输入,构建决策树评价模型,输出换流站直流设备健康状态评价结果。实验结果表明,该模型具备较好的多信息融合能力和样本均衡能力,且可从不同角度实现换流站直流设备健康状态评价,评价结果较为准确。

基于属性散射中心目标重构加权决策融合的SAR目标识别方法

针对合成孔径雷达(SAR)图像目标识别问题,采用原始图像及其属性散射中心目标重构结果进行决策融合。以核稀疏表示分类(KSRC)为基础分类器,对原始及重构SAR图像进行分类。KSRC通过引入核函数提升分类适应能力;目标重构可有效剔除原始SAR图像中的噪声成分。根据目标重构过程中重构结果与残差的能量关系评估原始SAR图像噪声水平,并以此为依据确定原始图像和重构图像决策结果的权重。采用加权融合手段对两个结果进行处理,判断测试样本的目标类别。基于MSTAR数据集对方法进行测试,实验结果证明了其有效性。