Clustering Of Objects Research Articles

Insider Threat Detection Based on Deep Clustering of Multi-Source Behavioral Events

With the continuous advancement of enterprise digitization, insider threats have become one of the primary cybersecurity concerns for organizations. Therefore, it is of great significance to develop an effective insider threat detection mechanism to ensure the security of enterprises. Most methods rely on artificial feature engineering and input the extracted user behavior features into a clustering-based unsupervised machine learning model for insider threat detection. However, feature extraction is independent of clustering-based unsupervised machine learning. As a result, user behavior features are not the most appropriate for clustering-based unsupervised machine learning, and thus, they reduce the insider threat detection accuracy. This paper proposes an insider threat detection method based on the deep clustering of multi-source behavioral events. On the one hand, the proposed method constructs an end-to-end deep clustering network and automatically learns the user behavior feature expression from multi-source behavioral event sequences. On the other hand, a deep clustering objective function is presented to jointly optimize the learning of feature representations and the clustering task for insider threat detection. This optimization can adjust the optimal user behavior features for the clustering model to improve the insider threat detection accuracy. The experimental results show that the proposed end-to-end insider threat detection model can accurately identify insider threats based on abnormal multi-source user behaviors in enterprise networks.

DFPC: Dynamic Fuzzy-based Primary User Aware clustering for Cognitive Radio Wireless Sensor Networks

Clustering-based routing solutions have proven to be efficient for wireless networks such as Wireless Sensor Networks (WSNs), Vehicular Ad Hoc Networks (VANETs), etc. Cognitive Radio WSN (CR-WSN) is a class of WSNs that consists of several resource-constrained Secondary Users (SUs), sink, and Primary Users (PUs). Compared to WSNs, there are several challenges in designing the clustering technique for CR-WSNs. As a result, one cannot directly apply the WSN clustering protocols to CR-WSNs. Developing a clustering protocol for CR-WSNs must address challenges such as ensuring PU protection, and SU connectivity, selecting the optimal Cluster Head (CH), and discovering the optimal cluster size. Present CR-WSN clustering solutions failed to resolve the trade-off among all essential clustering objectives. To address these challenges, this study presents a novel approach called Dynamic Fuzzy-based PU aware Clustering (DFPC) for CR-WSNs. DFPC uses a dynamic approach to discover the number of clusters, a fuzzy-based algorithm for optimal CH selection, and reliable multi-hop data transmission to ensure PU protection. To enhance the performance of CR-WSNs, an effective strategy was designed to define the optimal number of clusters using the network radius and live node. Fuzzy logic rules were formulated to assess the four CR-specific parameters for optimal CH selection. Finally, reliable intra- and intercluster data transmission routes are discovered to protect the PUs from malicious activities. The simulation results showed that the DFPC protocol achieved an improved average throughput of 48.04 and 46.49, a PDR of 93.36 and 84.37, and a reduced delay of 0.0271 and 0.0276 in static and dynamic topologies, respectively, which were better than those of ABCC, ATEEN, and LEACH protocols.

Cluster Strategy: Features of Sectoral and Intersectoral Legal Institutions Implementation

The paper discusses the processes of creation and functioning of cluster structures. Cluster approaches are very promising today: they allow for both creating new development institutions and managing the existing ones. The economic entities included in the cluster are also participants in the investment agreement at the same time. The author shows which subjective rights, absolute duties they have when united in an industrial-type cluster. The implementation of investment projects is mediated by the renewal of fixed assets. The innovation cluster then promises self-sufficiency, self-financing, but the creation of loan capital seems to be an extremely costly procedure. The institutions of the contract system that are able to stem the risks arising in this regard are considered. Various aspects of substitution transactions, in general, the interaction of investment and management contracts are summarized. The author notes that within the cluster, its beneficiaries by their actions contribute to both the accumulation of profits and the accumulation of losses. The more chances they have to pay off accounts receivable, the sooner they will turn their assets into liquid ones at full value. The functionality of cluster objects is brought to the fore, which outlines the shortest path to monetization of their development. The development of the Arctic resource potential (due to the cluster presumption) in this regard seems significant. The interaction of legal institutions is considered according to the concept of the so-called Triple helix, as far as it simplifies the creation of property complexes.

Scrutinising an inscrutable bark-nesting ant: Exploring cryptic diversity in the Rhopalomastix javana (Hymenoptera: Formicidae) complex using DNA barcodes, genome-wide MIG-seq and geometric morphometrics.

Overlooking cryptic species diversity has grave implications on assessments of climate change impacts on biodiversity, ecosystems and organismal populations. Discriminating between cryptic species has long been challenging even for seasoned taxonomists, as interspecies morphological differences are often indiscernible by visual observation. Multi-disciplinary methods involving genetic analyses in conjunction with quantitative morphological data, should therefore be used to investigate boundaries between cryptic species. We adopted an integrated approach combining analyses of mitochondrial COI barcodes, a genome-wide dataset obtained via multiplexed inter-simple sequence repeats (ISSRs) genotyping by sequencing (MIG-seq), and geometric morphometrics to investigate species divergences in the inscrutable Rhopalomastix javana species complex. Objective clustering of COI suggested five putative molecular species units divergent from each other by thresholds within 4.2-10.6% uncorrected pairwise distance. Phylogenetic analyses based on concatenated MIG-seq data also recovered and strongly supported the monophyly of five major lineages in agreement with COI clusters. Co-ancestry analyses based on MIG-seq data using fineRADstructure resolved variable patterns of admixture linked to geography, and potential genetic drift within some putative species. Geometric morphometric analyses of specimen images further detected statistically significant differences in at least one of three anatomical aspects (Head, Meso, Profile) between all pairs of putative species. Head shape (full-face view) was determined to be the most informative character for species diagnosis, with relatively high classification accuracy. Thin-plate spline deformation grids highlighted areas of high variation between species in each shape for deeper taxonomic scrutiny. The presence of species from multiple distinct lineages existing in near-sympatry firmly demonstrates that R. javana comprises more than one closely-related species, but exact species boundaries are difficult to ascertain. Differences in elevation and its associated abiotic effects on ant adaptations and reproductive phenology may contribute to restricting gene flow and maintaining species boundaries between sympatric populations of the R. javana complex. We further assess the advantages and limitations of geometric morphometrics as a taxonomic tool. Despite its drawbacks, our combined approach has helped draw important insights on cryptic diversity in R. javana, and also identified gaps of knowledge that await address. Results from this study will inform and prime future in-depth taxonomic investigation on the R. javana complex, including formal descriptions and establishment of the five putative species.

Atmospheric Circulation Patterns during the Summertime Precipitation in Southeastern Iran

The accurate recognition of atmospheric circulation patterns is vital for understanding the intricate relationships among various climatic elements. Therefore, the main goal of this study is to comprehensively identify circulation patterns during the occurrence of the summertime Extended Area Precipitation Event (EAPE) in southeastern Iran. The data used in this study encompass precipitation rates from synoptic and rain gauge stations, Geopotential Height (GPH), omega (upward motion speed), u-wind (east-west), and v-wind (north-south) components at different atmospheric levels, along with satellite images from the visible spectrum. In this research, both subjective and objective clustering methods have been utilized to identify synoptic circulation patterns based on 500-hPa GPH data. Summer precipitation was chosen for analysis because its characteristics and relationships with large-scale circulation patterns are less understood compared to those of winter precipitation. Examination of the 500-hPa GPH data for sixty-two identified cases of EAPE over southeast Iran revealed that the causative factors for these events are comprised of five recurring patterns (referred to here for convenience as AP, BP, CP, DP, and EP). Three of these patterns (AP, BP, and DP) significantly contributed to 71% of all EAPE cases. It was evident that the five patterns responsible for creating the EAPE in southeastern Iran had distinct directions.

Efficient Heart Disease Prediction Using PSO Optimized LSTM Classifier

Accurate diagnosis and prognosis of cardiovascular disease are essential in the medical field, enabling cardiologists to administer appropriate treatments. Machine learning has gained prominence in the healthcare sector due to its ability to identify patterns in data. Leveraging machine learning for classifying cardiovascular disease occurrences, significantly minimize misdiagnosis. This research focuses on developing a robust model capable of accurately predicting cardiovascular diseases, ultimately aiming to decrease the mortality rates associated with these conditions. To find clusters of data objects in a dataset, K-means clustering algorithm is implemented. The technique known as particle swarm optimization (PSO) is used to estimate solutions to extremely challenging or unsolvable numerical minimization and maximization problems. LSTM classifier, a deep learning technology is employed to categorize the attacks under various scenarios. In addition to greatly enhancing the healthcare system, this research gives medical practitioners access to an invaluable diagnostic tool. The outcomes of proposed work is examined using programs written in Python Jupyter software tool.

A between-cluster approach for clustering skew-symmetric data

In order to investigate exchanges between objects, a clustering model for skew-symmetric data is proposed, which relies on the between-cluster effects of the skew-symmetries that represent the imbalances of the observed exchanges between pairs of objects. The aim is to detect clusters of objects that share the same behaviour of exchange so that origin and destination clusters are identified. The proposed model is based on the decomposition of the skew-symmetric matrix pertaining to the imbalances between clusters into a sum of a number of off-diagonal block matrices. Each matrix can be approximated by a skew-symmetric matrix by using a truncated Singular Value Decomposition (SVD) which exploits the properties of the skew-symmetric matrices. The model is fitted in a least-squares framework and an efficient Alternating Least Squares algorithm is provided. Finally, in order to show the potentiality of the model and the features of the resulting clusters, an extensive simulation study and an illustrative application to real data are presented.

An approach to spatio-temporal analysis for climatic data

This work considers a large-scale multisite investigation of the effects of sunshine duration on Algerian territory. It also studies the correlation between sunshine duration and commonly used climatic parameters (temperature, water vapour pressure, evaporation, relative humidity and rainfall). Two data sets have served this purpose. The first one consists of sunshine duration measurements of the whole Algerian weather network over the period 1960-2002 while the second set provides climatic parameters collected in Dar El-Beida, Algiers (36° 41'N, 03°13' E) over the period 1950-2008. This station is the best in terms of fulfilling criteria such as long time data series (at least 50 years) as well as reliable measurements. To achieve the expected goal, we use an appropriate objective clustering method, named Principal Component Analysis (PCA) with a coupling of a Hierarchical Ascending Clustering (HAC) algorithm. PCA is often used not only for reducing the data before the actual clustering is carried out but also because it might help identifying the characteristics of the clusters. In this way, we establish a distribution of weather stations and identify the main homogeneous areas of the country that are more distinguishable and useful to our purpose as well as we classify the months of year. Obtained results divide Algeria into three distinct climate regions which daily monthly means of sunshine duration are also studied. Further information can be drawn through maps established by ‘MapInfo’ Software. According to the correlation matrix, it also turns out that sunshine duration strongly influences climatic parameters.

Objective and Subjective Clustering Methods for Verb Fluency Responses From Individuals With Alzheimer's Dementia and Cognitively Healthy Older Adults.

The verb fluency task has been researched using a variety of analysis methods and shown its sensitivity to declines in executive functioning and lexical retrieval abilities in various neurogenic populations. Few studies to date, however, have analyzed clusters and switches in the task, and there is a lack of robust analysis methods that preclude subjectivity and potential rater bias. The purpose of this study was to investigate the reliability when using subjective clustering methods and to determine the feasibility of using an objective clustering method to determine verb fluency performance in individuals with Alzheimer's dementia (IwDs) and cognitively healthy older adults (CHOAs). Responses from a verb fluency task were obtained from IwDs and CHOAs. Group differences were examined using an objective clustering method for multiple variables regarding clustering and switching. We also calculated the intrarater, interrater, and intermethod reliability using intraclass coefficients. Significant group differences were found when utilizing the objective clustering method in all variables except the average cluster size, with IwDs performing poorer than CHOAs. Intrarater reliability was excellent. Interrater reliability between two authors and intermethod reliability between the objective and subjective methods were variable ranging from moderate to good. The results from using the objective clustering method in this study are consistent with the previous literature, making it a viable option for clustering analyses on the verb fluency task, which naturally minimizes subjectivity and rater bias. Alternatively, employing a thoroughly validated and reliable subjective approach can also mitigate potential rater bias and improve replicability across studies. https://doi.org/10.23641/asha.24061017.

Residual Tuning: Toward Novel Category Discovery Without Labels.

Discovering novel visual categories from a set of unlabeled images is a crucial and essential capability for intelligent vision systems since it enables them to automatically learn new concepts with no need for human-annotated supervision anymore. To tackle this problem, existing approaches first pretrain a neural network with a set of labeled images and then fine-tune the network to cluster unlabeled images into a few categorical groups. However, their unified feature representation hits a tradeoff bottleneck between feature preservation on labeled data and feature adaptation on unlabeled data. To circumvent this bottleneck, we propose a residual-tuning approach, which estimates a new residual feature from the pretrained network and adds it with a previous basic feature to compute the clustering objective together. Our disentangled representation approach facilitates adjusting visual representations for unlabeled images and overcoming forgetting old knowledge acquired from labeled images, with no need of replaying the labeled images again. In addition, residual-tuning is an efficient solution, adding few parameters and consuming modest training time. Our results on three common benchmarks show consistent and considerable gains over other state-of-the-art methods, and further reduce the performance gap to the fully supervised learning setup. Moreover, we explore two extended scenarios, including using fewer labeled classes and continually discovering more unlabeled sets, where the results further signify the advantages and effectiveness of our residual-tuning approach against previous approaches. Our code is available at https://github.com/liuyudut/ResTune.

Black Holes in Globular Clusters

Black holes are the laboratory for extreme physics research in the universe, and the detection of black holes is one of the most cutting-edge international scientific goals at present. Globular clusters are old, self-gravity-bound object systems in the Milky Way that contain millions of stars with a high stellar densities inside. Hundreds of pulsar objects have been observed in globular clusters, but the existence of black holes in globular clusters remains an unsolved mystery. A globular cluster may have an intermediatemass black hole at its centre, with masses ranging from 1, 000 to 10, 000 times solar masses. At the same time, there may be a large number of stellar black holes in globular clusters, which will affect the evolution of globular clusters and form. some special objects in globular clusters, which can be observed and confirmed by astronomical observation equipment. This paper will focus on the possible observational features of the presence of black holes in globular clusters and their effects on the properties of globular clusters. Black holes are coupled with their host globular clusters through dynamics, which are important sites for the formation of black hole X-ray binaries and gravitational wave sources, and black holes change the dynamical structure and evolutionary processes of their host clusters. Therefore, searching for X-ray sources, detecting gravitational wave signals, and studying the characteristics of globular clusters are important astronomical evidence for detecting black holes in globular clusters.

Plasmatic thinking and tourism: Plasmatic modernity

This paper builds on assemblage theory to propose a new theoretical understanding of modernity. While the conceptual framing is meant for modernity at large, this paper locates its conceptual discussion in the context of tourism in Macao and illustrates how plasmatic thinking, the new conceptual framework proposed, advances analysis of aspiration, exploitation and freedom of its tour guides. Plasmatic thinking helps examinations of tourism labour to engage with the fragile and fluid nature of the sociomaterial environments. Instead of structures, networks or fluidities, plasmatic thinking sees the world as composed of ‘plasmas’ – ‘charged’ sociomaterial clustering of objects, humans and the processes between them. Plasmas are a form of charged matter falling outside solid, liquid and gaseous states and metaphorises the fragility and impermanence of sociomaterial situations for plasmas disintegrates when discharged. The attention to charges and fragility of plasmas helps describes both pandemic levels shocks and everyday disruptions. Through a plasmatic analysis of the falling apart and coming together of such plasmas and how they bring about significant consequences to Macao's tourism, I showcase plasmatic thinking as a theoretical approach which vividly uncovers the fragility and fluidity of modernity and the workings of power in our sociomaterial worlds.

Patch seriation to visualize data and model parameters

We developed a new seriation merit function for enhancing the visual information of data matrices. A local similarity matrix is calculated, where the average similarity of neighbouring objects is calculated in a limited variable space and a global function is constructed to maximize the local similarities and cluster them into patches by simple row and column ordering. The method identifies data clusters in a powerful way, if the similarity of objects is caused by some variables and these variables differ for the distinct clusters. The method can be used in the presence of missing data and also on more than two-dimensional data arrays. We show the feasibility of the method on different data sets: on QSAR, chemical, material science, food science, cheminformatics and environmental data in two- and three-dimensional cases. The method can be used during the development and the interpretation of artificial neural network models by seriating different features of the models. It helps to identify interpretable models by elucidating clusters of objects, variables and hidden layer neurons.Graphical

Eigenvector centrality in simplicial complexes of hypergraphs.

Hypergraph is the model of relations lying in clusters of objects. Identifying vital nodes is a fundamental problem in the analysis of the hypergraph. To reflect the multilayer feature of the hypergraph, in this paper, we deconstruct the hypergraph into a simplicial complex and analyze the homological dual relations of boundary and coboundary between simplices. For clarity, these two relations are summarized into a bidirectional graph, called the simplicial diagram, which provides a global framework for the exploration of the hypergraph. To determine the node importance in the hypergraph, we propose a parameter-free eigenvector centrality for weighted hypergraphs in terms of a simplicial complex, named Simplicial DualRank centrality. For each simplex, we define two indices of importance, the inner centrality and the outer centrality. Inner centrality transmits according to the relation of coboundary, which converts to outer centrality at the hyperlinks; in duality, outer centrality transmits according to the relation of boundary, which converts to inner centrality at the nodes. Therefore, a circuit of centrality is constructed on the simplicial diagram, the steady state of which defines the Simplicial DualRank centrality of all the simplices in the hypergraph. Moreover, we apply the Simplicial DualRank centrality to weighted complex networks, which results in a variant of the classical eigenvector centrality. Finally, experimental results in a science collaboration dataset show that the Simplicial DualRank can identify Nobel laureates from the prize-winning papers in Physics, top scientists should select collaborators more carefully to maintain their research quality, and scholars tend to find relatively effective collaborations in their future research.

Deep multi-view graph clustering network with weighting mechanism and collaborative training

With the development of graph convolutional network (GCN), which is powerful in graph embedding learning meanwhile can capture node feature information, deep multi-view graph clustering methods based on graph autoencoder have emerged as a new stream. Although they achieve satisfactory performance, they still have the following weaknesses: (1) some of them model the weights for different views in the encoder part by using attention in the multi-view embedding fusion layer, but fail to consider the weighting in the decoder part to measure the contributions of different views to the reconstruction loss. (2) Most of them directly conduct clustering on the multi-view common embedding layer, but fail to guarantee the alignment of clustering results of different views. To this end, we propose a novel GCN-based deep multi-view graph clustering network with weighting mechanism and collaborative training (DMVGC). The model is composed of multiple view-specific graph encoders and a unified graph decoder. Besides the specially-designed attention module in the encoder part, we construct a reconstruction loss with adaptive weighting mechanism in the decoder part. Additionally, a collaborative self-training clustering objective is jointly conducted on each view-specific embedding layer and the common embedding layer, to make the embedding of each view clustering-friendly toward a common partition. Experiments on several datasets demonstrate the effectiveness of our model.

Implicit models, latent compression, intrinsic biases, and cheap lunches in community detection.

The task of community detection, which aims to partition a network into clusters of nodes to summarize its large-scale structure, has spawned the development of many competing algorithms with varying objectives. Some community detection methods are inferential, explicitly deriving the clustering objective through a probabilistic generative model, while other methods are descriptive, dividing a network according to an objective motivated by a particular application, making it challenging to compare these methods on the same scale. Here we present a solution to this problem that associates any community detection objective, inferential or descriptive, with its corresponding implicit network generative model. This allows us to compute the description length of a network and its partition under arbitrary objectives, providing a principled measure to compare the performance of different algorithms without the need for "ground-truth" labels. Our approach also gives access to instances of the community detection problem that are optimal to any given algorithm and in this way reveals intrinsic biases in popular descriptive methods, explaining their tendency to overfit. Using our framework, we compare a number of community detection methods on artificial networks and on a corpus of over 500 structurally diverse empirical networks. We find that more expressive community detection methods exhibit consistently superior compression performance on structured data instances, without having degraded performance on a minority of situations where more specialized algorithms perform optimally. Our results undermine the implications of the "no free lunch" theorem for community detection, both conceptually and in practice, since it is confined to unstructured data instances, unlike relevant community detection problems which are structured by requirement.

Instance segmentation of stack composed of unknown objects

The article reviews neural network architectures designed for the segmentation task. It focuses mainly on instance segmentation of stacked objects. The main assumption is that segmentation is based on a color image with an additional depth layer. The paper also introduces the Stacked Bricks Dataset based on three cameras: RealSense L515, ZED2, and a synthetic one. Selected architectures: DeepLab, Mask RCNN, DEtection TRansformer, Geometry-Aware Instance Segmentation, Shapemask, Synthetic Depth Mask RCNN, Synthetic Fusion Mask RCNN (SF-Mask), Unseen Object Instance Segmentation (UOIS), Unseen Object Clustering (UOC), and You Look Only At Coefficients, have been tested on various datasets. The results show that the best architectures for stacked elements segmentation are UOIS, SF-Mask, and UOC.

An Optimization-Based Order-and-Cut Approach for Fair Clustering of Data Sets

Machine learning algorithms have been increasingly integrated into applications that significantly affect human lives. This surged an interest in designing algorithms that train machine learning models to minimize training error and imposing a certain level of fairness. In this paper, we consider the problem of fair clustering of data sets. In particular, given a set of items each associated with a vector of nonsensitive attribute values and a categorical sensitive attribute (e.g., gender, race, etc.), our goal is to find a clustering of the items that minimizes the loss (i.e., clustering objective) function and imposes fairness measured by Rényi correlation. We propose an efficient and scalable in-processing algorithm, driven by findings from the field of combinatorial optimization, that heuristically solves the underlying optimization problem and allows for regulating the trade-off between clustering quality and fairness. The approach does not restrict the analysis to a specific loss function, but instead considers a more general form that satisfies certain desirable properties. This broadens the scope of the algorithm’s applicability. We demonstrate the effectiveness of the algorithm for the specific case of k-means clustering as it is one of the most extensively studied and widely adopted clustering schemes. Our numerical experiments reveal the proposed algorithm significantly outperforms existing methods by providing a more effective mechanism to regulate the trade-off between loss and fairness. History: Rema Padman served as the senior editor for this article. Data Ethics & Reproducibility Note: The code capsule is available on Code Ocean at https://codeocean.com/capsule/9556728/tree and in the e-Companion to the this article (available at https://doi.org/10.1287/ijds.2022.0005 ).

Active learning with fairness-aware clustering for fair classification considering multiple sensitive attributes

Fairness concerns have recently been gaining increasing attention in machine learning (ML) research and applications. ML models typically require massive data, which can be costly and challenging for collecting and labeling. To address these concerns, active learning has been proposed as a promising approach to build accurate and fair ML models by interactively querying an oracle within a labeling budget. However, current research on fair active learning only focuses on single sensitive attribute, where the model often show unfairness on subgroups defined by the intersections of different protected attribute values. In this article, a novel active learning approach with a fairness-aware clustering technique is proposed for fair classification that can simultaneously ensure the equality of fairness metrics across multiple sensitive attributes. The fair clustering method enables to control the trade-off levels between the clustering and fairness objectives with multiple sensitive attributes, which mitigates bias and protects the unprivileged classes by balancing their representativeness in all clusters. A tight upper bound is theoretically derived for our fairness objective based on its concave-convex decomposition, which can be jointly optimized with the clustering objective through Majorize-Minimize technique and enable parallel computing. The active learning sampler aims to select highly informative and representative samples from fair clusters to enhance the model classification accuracy and fairness. The effectiveness of our proposed approach is demonstrated by comparing with state-of-the-art methods on real-world datasets. The experimental results are accordant with theoretic analysis, and show that (i) our approach significantly improves the fairness while maintaining model classification accuracy; (ii) our approach outperforms state-of-the-art fair active learning methods across fairness metrics on both single and multiple sensitive attributes for fair classification tasks.

Methods for the Formation of Spatiotemporal Clusters of Objects in an Unfriendly Environment

Methods for the Formation of Spatiotemporal Clusters of Objects in an Unfriendly Environment