Model Change Active Learning in Graph-Based Semi-supervised Learning

Active learning in semi-supervised classification involves introducing additional labels for unlabelled data to improve the accuracy of the underlying classifier.A challenge is to identify which points to label to best improve performance while limiting the number of new labels."Model Change"active learning quantifies the resulting change incurred in the classifier by introducing the additional label(s).We pair this idea with graph-based semi-supervised learning(SSL)methods,that use the spectrum of the graph Laplacian matrix,which can be truncated to avoid prohibitively large computational and storage costs.We consider a family of convex loss functions for which the acquisition function can be efficiently approximated using the Laplace approximation of the posterior distribution.We show a variety of multiclass examples that illustrate improved performance over prior state-of-art.

Multi-resolution graph-based clustering analysis for lithofacies identifi cation from well log data: Case study of intraplatform bank gas fi elds, Amu Darya Basin

In this study, we used the multi-resolution graph-based clustering （MRGC） method for determining the electrofacies （EF） and lithofacies （LF） from well log data obtained from the intraplatform bank gas fields located in the Amu Darya Basin. The MRGC could automatically determine the optimal number of clusters without prior knowledge about the structure or cluster numbers of the analyzed data set and allowed the users to control the level of detail actually needed to define the EF. Based on the LF identification and successful EF calibration using core data, an MRGC EF partition model including five clusters and a quantitative LF interpretation chart were constructed. The EF clusters 1 to 5 were interpreted as lagoon, anhydrite flat, interbank, low-energy bank, and high-energy bank, and the coincidence rate in the cored interval could reach 85%. We concluded that the MRGC could be accurately applied to predict the LF in non-cored but logged wells. Therefore, continuous EF clusters were partitioned and corresponding LF were characteristics ＆different LF were analyzed interpreted, and the distribution and petrophysical in the framework of sequence stratigraphy.

BotSward: Centrality Measures for Graph-Based Bot Detection Using Machine Learning

The number of botnet malware attacks on Internet devices has grown at an equivalent rate to the number of Internet devices that are connected to the Internet.Bot detection using machine learning(ML)with flow-based features has been extensively studied in the literature.Existing flow-based detection methods involve significant computational overhead that does not completely capture network communication patterns that might reveal other features ofmalicious hosts.Recently,Graph-Based Bot Detection methods using ML have gained attention to overcome these limitations,as graphs provide a real representation of network communications.The purpose of this study is to build a botnet malware detection system utilizing centrality measures for graph-based botnet detection and ML.We propose BotSward,a graph-based bot detection system that is based on ML.We apply the efficient centrality measures,which are Closeness Centrality(CC),Degree Centrality(CC),and PageRank(PR),and compare them with others used in the state-of-the-art.The efficiency of the proposed method is verified on the available Czech Technical University 13 dataset(CTU-13).The CTU-13 dataset contains 13 real botnet traffic scenarios that are connected to a command-and-control(C&C)channel and that cause malicious actions such as phishing,distributed denial-of-service(DDoS)attacks,spam attacks,etc.BotSward is robust to zero-day attacks,suitable for large-scale datasets,and is intended to produce better accuracy than state-of-the-art techniques.The proposed BotSward solution achieved 99%accuracy in botnet attack detection with a false positive rate as low as 0.0001%.

A Novel Method for Node Connectivity with Adaptive Dragonfly Algorithm and Graph-Based m-Connection Establishment in MANET

Maximizing network lifetime is measured as the primary issue in Mobile Ad-hoc Networks(MANETs).In geographically routing based models,packet transmission seems to be more appropriate in dense circumstances.The involvement of the Heuristic model directly is not appropriate to offer an effectual solution as it becomes NP-hard issues;therefore investigators concentrate on using Meta-heuristic approaches.Dragonfly Optimization(DFO)is an effective meta-heuristic approach to resolve these problems by providing optimal solutions.Moreover,Meta-heuristic approaches(DFO)turn to be slower in convergence problems and need proper computational time while expanding network size.Thus,DFO is adaptively improved as Adaptive Dragonfly Optimization(ADFO)to fit this model and re-formulated using graph-based m-connection establishment(G-𝑚𝑚CE)to overcome computational time and DFO’s convergence based problems,considerably enhancing DFO performance.In(G-𝑚𝑚CE),Connectivity Zone(CZ)is chosen among source to destination in which optimality should be under those connected regions and ADFO is used for effective route establishment in CZ indeed of complete networking model.To measure complementary features of ADFO and(G-𝑚𝑚CE),hybridization of DFO-(G-𝑚𝑚CE)is anticipated over dense circumstances with reduced energy consumption and delay to enhance network lifetime.The simulation was performed in MATLAB environment.

Graph-Based Replication and Two Factor Authentication in Cloud Computing

Many cutting-edge methods are now possible in real-time commercial settings and are growing in popularity on cloud platforms.By incorporating new,cutting-edge technologies to a larger extent without using more infrastructures,the information technology platform is anticipating a completely new level of devel-opment.The following concepts are proposed in this research paper:1)A reliable authentication method Data replication that is optimised;graph-based data encryp-tion and packing colouring in Redundant Array of Independent Disks(RAID)sto-rage.At the data centre,data is encrypted using crypto keys called Key Streams.These keys are produced using the packing colouring method in the web graph’s jump graph.In order to achieve space efficiency,the replication is carried out on optimised many servers employing packing colours.It would be thought that more connections would provide better authentication.This study provides an innovative architecture with robust security,enhanced authentication,and low cost.

A graph-based pan-genome of Brassica oleracea provides new insights into its domestication and morphotype diversification

The domestication of Brassica oleracea has resulted in diverse morphological types with distinct patterns of organ development.Here we report a graph-based pan-genome of B.oleracea constructed from high-quality genome assemblies of different morphotypes.The pan-genome harbors over 200 structural variant hotspot regions enriched in auxin-andflowering-related genes.Population genomic analyses revealed that early domestication of B.oleracea focused on leaf or stem development.Geneflows resulting from agricultural practices and variety improvement were detected among different morphotypes.Selective-sweep and pan-genome analyses identified an auxin-responsive small auxin up-regulated RNA gene and a CLAV-ATA3/ESR-RELATED family gene as crucial players in leaf–stem differentiation during the early stage of B.oleracea domestication and the BoKAN1 gene as instrumental in shaping the leafy heads of cabbage and Brussels sprouts.Our pan-genome and functional analyses further revealed that variations in the BoFLC2 gene play key roles in the divergence of vernalization andflowering characteristics among different morphotypes,and variations in thefirst intron of BoFLC3 are involved infine-tuning theflowering process in cauliflower.This study provides a comprehensive understanding of the pan-genome of B.oleracea and sheds light on the domestication and differential organ development of this globally important crop species.

基于超像素的Graph-Based图像分割算法

针对EGBIS分割算法中的过分割问题,提出了一种基于超像素的graph-based图像分割算法SGBIS.首先,对图像进行基于简单线性迭代聚类（SLIC）的超像素预分割;然后以每个超像素作为节点构造带权无向图,以相邻超像素颜色平均值的欧式距离作为图中边的权值;最后利用基于图的算法合并超像素得到分割结果.用VI、PRI和F值3个指标分析了算法性能,结果表明,新算法可以得到更为理想的分割效果;引入交互分割区域合并,也可满足用户图像分割的需求.

Instance selection method for improving graph-based semi-supervised learning

Graph-based semi-supervised learning is an important semi-supervised learning paradigm. Although graphbased semi-supervised learning methods have been shown to be helpful in various situations, they may adversely affect performance when using unlabeled data. In this paper, we propose a new graph-based semi-supervised learning method based on instance selection in order to reduce the chances of performance degeneration. Our basic idea is that given a set of unlabeled instances, it is not the best approach to exploit all the unlabeled instances; instead, we should exploit the unlabeled instances that are highly likely to help improve the performance, while not taking into account the ones with high risk. We develop both transductive and inductive variants of our method. Experiments on a broad range of data sets show that the chances of performance degeneration of our proposed method are much smaller than those of many state-of-the-art graph-based semi-supervised learning methods.

A graph-based approach for the structural analysis of road and building layouts

A better understanding of the relationship between the structure and functions of urban and suburban spaces is one of the avenues of research still open for geographical information science.The research presented in this paper develops several graph-based metrics whose objective is to characterize some local and global structural properties that reflect the way the overall building layout can be cross-related to the one of the road layout.Such structural properties are modeled as an aggregation of parcels,buildings,and road networks.We introduce several computational measures(Ratio Minimum Distance,Minimum Ratio Minimum Distance,and Metric Compactness)that respectively evaluate the capability for a given road to be connected with the whole road network.These measures reveal emerging sub-network structures and point out differences between less-connective and moreconnective parts of the network.Based on these local and global properties derived from the topological and graph-based representation,and on building density metrics,this paper proposes an analysis of road and building layouts at different levels of granularity.The metrics developed are applied to a case study in which the derived properties reveal coherent as well as incoherent neighborhoods that illustrate the potential of the approach and the way buildings and roads can be relatively connected in a given urban environment.Overall,and by integrating the parcels and buildings layouts,this approach complements other previous and related works that mainly retain the configurational structure of the urban network as well as morphological studies whose focus is generally limited to the analysis of the building layout.

Low Data Overlab Rate Graph-Based SLAM with Distributed Submap Strategy

Simultaneous localization and mapping(SLAM)is widely used in many robot applications to acquire the unknown environment's map and the robots location.Graph-based SLAM is demonstrated to be effective in large-scale scenarios,and it intuitively performs the SLAM as a pose graph.But because of the high data overlap rate,traditional graph-based SLAM is not efficient in some respects,such as real time performance and memory usage.To reduce1 data overlap rate,a graph-based SLAM with distributed submap strategy(DSS)is presented.In its front-end,submap based scan matching is processed and loop closing detection is conducted.Moreover in its back-end,pose graph is updated for global optimization and submap merging.From a series of experiments,it is demonstrated that graph-based SLAM with DSS reduces 51.79%data overlap rate,decreases 39.70%runtime and 24.60%memory usage.The advantages over other low overlap rate method is also proved in runtime,memory usage,accuracy and robustness performance.

Graph-based robot optimal path planning with bio-inspired algorithms

Recently,bio-inspired algorithms have been increasingly explored for autonomous robot path planning on grid-based maps.However,these approaches endure performance degradation as problem complexity increases,often resulting in lengthy search times to find an optimal solution.This limitation is particularly critical for real-world applications like autonomous off-road vehicles,where highquality path computation is essential for energy efficiency.To address these challenges,this paper proposes a new graph-based optimal path planning approach that leverages a sort of bio-inspired algorithm,improved seagull optimization algorithm(iSOA)for rapid path planning of autonomous robots.A modified Douglas–Peucker(mDP)algorithm is developed to approximate irregular obstacles as polygonal obstacles based on the environment image in rough terrains.The resulting mDPderived graph is then modeled using a Maklink graph theory.By applying the iSOA approach,the trajectory of an autonomous robot in the workspace is optimized.Additionally,a Bezier-curve-based smoothing approach is developed to generate safer and smoother trajectories while adhering to curvature constraints.The proposed model is validated through simulated experiments undertaken in various real-world settings,and its performance is compared with state-of-the-art algorithms.The experimental results demonstrate that the proposed model outperforms existing approaches in terms of time cost and path length.

Graph-based semi-supervised learning

The recent years have witnessed a surge of interests in graph-based semi-supervised learning(GBSSL).In this paper,we will introduce a series of works done by our group on this topic including:1)a method called linear neighborhood propagation(LNP)which can automatically construct the optimal graph;2)a novel multilevel scheme to make our algorithm scalable for large data sets;3)a generalized point charge scheme for GBSSL;4)a multilabel GBSSL method by solving a Sylvester equation;5)an information fusion framework for GBSSL;and 6)an application of GBSSL on fMRI image segmentation.

A graph-based two-stage classification network for mobile screen defect inspection

Defect inspection,also known as defect detection,is significant in mobile screen quality control.There are some challenging issues brought by the characteristics of screen defects,including the following:(1)the problem of interclass similarity and intraclass variation,(2)the difficulty in distinguishing low contrast,tiny-sized,or incomplete defects,and(3)the modeling of category dependencies for multi-label images.To solve these problems,a graph reasoning module,stacked on a classification module,is proposed to expand the feature dimension and improve low-quality image features by exploiting category-wise dependency,image-wise relations,and interactions between them.To further improve the classification performance,the classifier of the classification module is redesigned as a cosine similarity function.With the help of contrastive learning,the classification module can better initialize the category-wise graph of the reasoning module.Experiments on the mobile screen defect dataset show that our two-stage network achieves the following best performances:97.7%accuracy and 97.3%F-measure.This proves that the proposed approach is effective in industrial applications.

改进的一种图论分割方法在舌像分割中的应用

由于人舌体的特殊性质,从舌像中直接分割舌体时常存在过分割和过合并现象。针对舌像特点提出了一种结合图论分割和多分辨率分割的图像分割算法,用一种图论分割算法让舌像在两种分辨率下分别进行分割,根据两种分割的结果把它们进行交或者并处理,从而有效地分割出舌体。实验结果表明这种方法能够有效避免直接使用图论分割时出现分割过度或者欠分割的情况。

采用相干斑抑制图割的SAR海冰图像分割

针对SAR海冰图像分割受相干斑噪声干扰严重的问题,在MRF框架下,提出一种分割新算法—SRGB-RMRF。算法首先根据相干斑噪声统计特性,对传统graph-based方法的梯度和区域内部差异计算公式重新定义,得到适用于SAR图像的相干斑抑制graph-based(SRGB)初始分割新方法。其次,结合区域间强度差异,在SRGB方法得到的区域邻接图上构建区域MRF模型。在合成SAR海冰图像和真实SAR海冰图像上的实验结果表明,与现有区域MRF算法相比,SRGB-RMRF算法能够实现更为准确的SAR海冰图像分割。

Mobile robot localization algorithm based on multi-sensor information fusion

In order to effectively reduce the uncertainty error of mobile robot localization with a single sensor and improve the accuracy and robustness of robot localization and mapping,a mobile robot localization algorithm based on multi-sensor information fusion(MSIF)was proposed.In this paper,simultaneous localization and mapping(SLAM)was realized on the basis of laser Rao-Blackwellized particle filter(RBPF)-SLAM algorithm and graph-based optimization theory was used to constrain and optimize the pose estimation results of Monte Carlo localization.The feature point extraction and quadrilateral closed loop matching algorithm based on oriented FAST and rotated BRIEF(ORB)were improved aiming at the problems of generous calculation and low tracking accuracy in visual information processing by means of the three-dimensional(3D)point feature in binocular visual reconstruction environment.Factor graph model was used for the information fusion under the maximum posterior probability criterion for laser RBPF-SLAM localization and binocular visual localization.The results of simulation and experiment indicate that localization accuracy of the above-mentioned method is higher than that of traditional RBPF-SLAM algorithm and general improved algorithms,and the effectiveness and usefulness of the proposed method are verified.

Conceptual Modular Design of Auto Body Frame Based on Hybrid Optimization Method

This article presents a systematic research methodology of modular design for conceptual auto body frame by hybrid optimization method.A modified graph-based decomposition optimization algorithm is utilized to generate an optimal BIW assembly topo model composed of“potential modules”.The consistency constraint function in collaborative optimization is extended to maximize the commonality of modules and minimize the performance loss of all car types in the same product family simultaneously.A novel screening method is employed to select both“basic structures”and“reinforcement”modules based on the dimension optimization of the manufacturing elements and the optimal assembly mode;this allows for a more exhaustive modular platform design in contrast with existing methods.The proposed methodology is applied to a case study for the modular design of three conceptual auto body types in the same platform to validate its feasibility and effectiveness.

A Unique Discrete Wavelet&Deterministic Walk-Based Glaucoma Classification Approach Using Image-Specific Enhanced Retinal Images

Glaucoma is a group of ocular atrophy diseases that cause progressive vision loss by affecting the optic nerve.Because of its asymptomatic nature,glaucoma has become the leading cause of human blindness worldwide.In this paper,a novel computer-aided diagnosis(CAD)approach for glaucomatous retinal image classification has been introduced.It extracts graph-based texture features from structurally improved fundus images using discrete wavelet-transformation(DWT)and deterministic tree-walk(DTW)procedures.Retinal images are considered from both public repositories and eye hospitals.Images are enhanced with image-specific luminance and gradient transitions for both contrast and texture improvement.The enhanced images are mapped into undirected graphs using DTW trajectories formed by the image’s wavelet coefficients.Graph-based features are extracted fromthese graphs to capture image texture patterns.Machine learning(ML)classifiers use these features to label retinal images.This approach has attained an accuracy range of 93.5%to 100%,82.1%to 99.3%,95.4%to 100%,83.3%to 96.6%,77.7%to 88.8%,and 91.4%to 100%on the ACRIMA,ORIGA,RIM-ONE,Drishti,HRF,and HOSPITAL datasets,respectively.The major strength of this approach is texture pattern identification using various topological graphs.It has achieved optimal performance with SVM and RF classifiers using biorthogonal DWT combinations on both public and patients’fundus datasets.The classification performance of the DWT-DTW approach is on par with the contemporary state-of-the-art methods,which can be helpful for ophthalmologists in glaucoma screening.

Text Augmentation-Based Model for Emotion Recognition Using Transformers

Emotion Recognition in Conversations(ERC)is fundamental in creating emotionally intelligentmachines.Graph-BasedNetwork(GBN)models have gained popularity in detecting conversational contexts for ERC tasks.However,their limited ability to collect and acquire contextual information hinders their effectiveness.We propose a Text Augmentation-based computational model for recognizing emotions using transformers(TA-MERT)to address this.The proposed model uses the Multimodal Emotion Lines Dataset(MELD),which ensures a balanced representation for recognizing human emotions.Themodel used text augmentation techniques to producemore training data,improving the proposed model’s accuracy.Transformer encoders train the deep neural network(DNN)model,especially Bidirectional Encoder(BE)representations that capture both forward and backward contextual information.This integration improves the accuracy and robustness of the proposed model.Furthermore,we present a method for balancing the training dataset by creating enhanced samples from the original dataset.By balancing the dataset across all emotion categories,we can lessen the adverse effects of data imbalance on the accuracy of the proposed model.Experimental results on the MELD dataset show that TA-MERT outperforms earlier methods,achieving a weighted F1 score of 62.60%and an accuracy of 64.36%.Overall,the proposed TA-MERT model solves the GBN models’weaknesses in obtaining contextual data for ERC.TA-MERT model recognizes human emotions more accurately by employing text augmentation and transformer-based encoding.The balanced dataset and the additional training samples also enhance its resilience.These findings highlight the significance of transformer-based approaches for special emotion recognition in conversations.

Geographic Drone-based Route Optimization Approach for Emergency Area Ad-Hoc Network

Wireless sensor Mobile ad hoc networks have excellent potential in moving and monitoring disaster area networks on real-time basis.The recent challenges faced in Mobile Ad Hoc Networks(MANETs)include scalability,localization,heterogeneous network,self-organization,and self-sufficient operation.In this background,the current study focuses on specially-designed communication link establishment for high connection stability of wireless mobile sensor networks,especially in disaster area network.Existing protocols focus on location-dependent communications and use networks based on typically-used Internet Protocol(IP)architecture.However,IP-based communications have a few limitations such as inefficient bandwidth utilization,high processing,less transfer speeds,and excessive memory intake.To overcome these challenges,the number of neighbors(Node Density)is minimized and high Mobility Nodes(Node Speed)are avoided.The proposed Geographic Drone Based Route Optimization(GDRO)method reduces the entire overhead to a considerable level in an efficient manner and significantly improves the overall performance by identifying the disaster region.This drone communicates with anchor node periodically and shares the information to it so as to introduce a drone-based disaster network in an area.Geographic routing is a promising approach to enhance the routing efficiency in MANET.This algorithm helps in reaching the anchor(target)node with the help of Geographical Graph-Based Mapping(GGM).Global Positioning System(GPS)is enabled on mobile network of the anchor node which regularly broadcasts its location information that helps in finding the location.In first step,the node searches for local and remote anticipated Expected Transmission Count(ETX),thereby calculating the estimated distance.Received Signal Strength Indicator(RSSI)results are stored in the local memory of the node.Then,the node calculates the least remote anticipated ETX,Link Loss Rate,and information to the new location.Freeway Heuristic algorithm improves the data speed,efficiency and determines the path and optimization problem.In comparison with other models,the proposed method yielded an efficient communication,increased the throughput,and reduced the end-to-end delay,energy consumption and packet loss performance in disaster area networks.