Effective Clustering Approach Research Articles

Applying a two-step cluster algorithm in traffic accident data analysis

Cluster analysis is often employed as the initial stage in organizing heterogeneous data into homogeneous groups. Choosing an effective clustering approach and an ideal number of clusters in a traffic accident dataset might be complex and challenging. This study aims to evaluate the effectiveness of k-means and two-step cluster methods. Subsequently, the two-step cluster method and GIS are applied to analyze the traffic accident datasets from 2015 to 2017 in Hanoi, Vietnam. First, according to the Silhouette score, the two-step cluster method achieved a higher score of 0.563, while the k-means method scored 0.341. A higher Silhouette score indicates more well-defined clusters. Second, the research suggests combining the two-step cluster method with GIS for analyzing traffic accident datasets. The outcome identifies five typical types of accidents in Hanoi. In addition, the locations of various accident types were visually illustrated on a map, enabling traffic officials to recommend precise and urgent countermeasures. Importantly, the clustering results reveal that the two-step cluster method exhibits a significantly higher rate of homogeneous data in the clusters compared to the k-means method. This study demonstrates that the two-step cluster method is not only more effective than the k-means method in terms of clustering ability but also in data pre-processing. The study's results enable authorities to gain a more detailed understanding of typical traffic accident patterns in Hanoi. Besides, the employed methods could potentially be applied to other regions, providing an additional avenue for analysis

Cluster stability-driven optimization for enhanced routing in heterogeneous vehicular networks

The new era of the Internet of Things is promoting the evolution of self-driving vehicles into connected and autonomous vehicles (CAVs). The deployment of CAVs in smart cities is highly dependent on the performance of their underlying networks known as vehicular networks. Designing an effective clustering approach is of great importance in such dynamic networks as it can significantly improve the reliability and scalability of the routing protocols. In this paper, we consider a heterogeneous vehicular network architecture that supports vehicle-to-vehicle and vehicle-to-infrastructure communications based on IEEE 802.11p and cellular networks (LTE/5G) with direct communications, namely cellular vehicle-to-everything (C-V2X) technologies. We introduce a novel clustering scheme for real-time routing based on the proposed network architecture. We formulate the problem of optimal clustering for connected and autonomous vehicles (OCCAV) and show that the problem is NP-hard. We then develop a clusters' stability maximization algorithm (CSM), which utilizes the stability degree of vehicles over a prediction horizon to efficiently solve the optimization problem in real-time. The algorithm is used within a rolling horizon framework for continuously solving the problem, making the resulting clusters adaptive to future traffic dynamics. We propose a hybrid routing protocol based on our clustering scheme, aiming to improve the packet delivery ratio and reduce the average delivery delay. For evaluation purposes, we use extensive realistic simulations based on mobility scenarios validated using real vehicular trajectories. The results demonstrate that our clustering scheme improves the alternative algorithms in terms of the average cluster head duration, cluster head change rate, cluster member duration, and overall stability by 61%, 62%, 44%, and 52%, respectively. CSM also outperforms clustering overhead by 54%. Compared to the other cluster-based routing, our scheme also achieves a higher packet delivery ratio by up to 30%, and a lower average delay by up to 52%. To provide an in-depth analysis of the optimality of our scheme and its alternatives, we also use the Gurobi optimizer to find an optimal solution to the OCCAV problem. The results suggest that our scheme can achieve near-optimal cluster stability.

Three-way evidence theory-based density peak clustering with the principle of justifiable granularity

Clustering by fast search and find of density peaks (DPC) is an effective clustering approach that can find all the cluster centers at once with just one parameter and without iterative processing. However, the cutoff distance, a key parameter of density measurement in the DPC approach, affects the quality of the final clustering results. Its selection relies on experimental experience and lacks of a semantic explanation. Furthermore, the allocation strategy of the traditional DPC approach may cause several points to be assigned incorrectly, leading to subsequent points being assigned incorrectly and ultimately forming continuous allocation errors. To overcome the deficiencies, this paper proposes a novel three-way evidence theory-based density peak clustering with the principle of justifiable granularity (3 W-PEDP). First, the computation of the cutoff distance is converted into the search for nearest neighbors. From the perspective of granular computing, 3 W-PEDP transforms the neighbor selection issue into the construction of justifiable granularity. And the optimal neighbors can be achieved with the construction of coverage and specificity criteria. Second, inspired by three-way clustering, we adopt a two-stage method for sample allocation. On the one hand, for core point allocation, a two-layer nearest neighbor is constructed based on the achieved optimal neighbors. On the other hand, we designed a new evidence mass function to guide us in assigning the remaining points. In this novel evidence mass function, not only the labels of the assigned samples are considered, but also the information of the neighborhoods around the unassigned samples is fused. Finally, we assess the effectiveness of 3 W-PEDP on numerous public synthetic datasets and UCI real-world datasets. Then, detail comparing results with several popular clustering methods are presented. In addition, experimental studies verify the effectiveness of constructing justifiable granularity in selecting the optimal neighbors. The experimental results demonstrate 3 W-PEDP has good adaptability and robustness, which can achieve better clustering performance. Our source code is available at https://github.com/Luyangabc/3W-PEDP.

A new and effective two-step clustering approach for single cell RNA sequencing data

BackgroundThe rapid devolvement of single cell RNA sequencing (scRNA-seq) technology leads to huge amounts of scRNA-seq data, which greatly advance the research of many biomedical fields involving tissue heterogeneity, pathogenesis of disease and drug resistance etc. One major task in scRNA-seq data analysis is to cluster cells in terms of their expression characteristics. Up to now, a number of methods have been proposed to infer cell clusters, yet there is still much space to improve their performance.ResultsIn this paper, we develop a new two-step clustering approach to effectively cluster scRNA-seq data, which is called TSC — the abbreviation of Two-Step Clustering. Particularly, by dividing all cells into two types: core cells (those possibly lying around the centers of clusters) and non-core cells (those locating in the boundary areas of clusters), we first clusters the core cells by hierarchical clustering (the first step) and then assigns the non-core cells to the corresponding nearest clusters (the second step). Extensive experiments on 12 real scRNA-seq datasets show that TSC outperforms the state of the art methods.ConclusionTSC is an effective clustering method due to its two-steps clustering strategy, and it is a useful tool for scRNA-seq data analysis.

Experimental Analysis of Fuzzy Clustering Techniques for Outlier Detection

Fuzzy clustering is an effective clustering approach that divides the dataset into fuzzy segments. Fuzzy clustering and outlier detection are two interconnected processes. Outlier detection is essential because it discloses hidden patterns and crucial information about a dataset and is beneficial in a wide range of applications like fraud detection, military surveillance, identifying computer network intrusions, image processing, insurance or health care, etc. Many fuzzy clustering techniques are robust to anomalies since they limit the outlier's impact on the cluster's centroid.In this paper, we have compared four fuzzy clustering techniques namely Fuzzy C-means (FCM), Noise Clustering (NC), Credibilistic Fuzzy C-means (CFCM), and Density Oriented Fuzzy C - Means (DOFCM) on the basis of some crucial properties that are essential for efficient outlier detection. To better evaluate the feasibility of all these techniques, we experimentally evaluated these techniques on six datasets (two Real and four Synthetic) in the presence of noise and outliers. The results of the comparative analysis show that DOFCM outperforms all other techniques in terms of outlier detection and cluster formation. This research will surely benefit any researcher willing to detect outliers in their study using fuzzy clustering or either implicitly or explicitly working on fuzzy clusters.

Secure PD-NOMA with Multi-User Cooperation and User Clustering in Both Uplink and Downlink PD-NOMA

The power domain non-orthogonal multiple access (PD-NOMA) scheme has gained tremendous interest with the multiple access behavior for fifth-generation (5G) wireless communication. Although the overall performance is improved through accurate power distribution among users’ signals, it depends on the user clustering strategy. Moreover, the PD-NOMA communication is not completely secured due to its broadcast nature, which is still a major problem. This paper presents a novel low-complexity short code-based technique utilized by the registered users and the 5G base station (gNodeB) for communication. By doing so, the PD-NOMA scheme is made secure from unregistered users or eavesdroppers. We proposed a three-step user clustering strategy that selects the best cluster among all the possible clusters to improve the overall performance. The proposed clustering strategy achieves a low outage probability in PD-NOMA systems. Moreover, it uses a multi-user decode and forward cooperative relaying scheme with PD-NOMA (Cop-PD-NOMA) to increase the coverage range of the gNodeB. In the multi-user Cop-PD-NOMA, the strong users (near users) are used as relay stations to aid the weak users (far users) by the decode and forward (D&F) technique. The proposed work provides a secure PD-NOMA network and the most effective user clustering approach during validation. The bit-error-rate (BER) comparisons demonstrate that multi-user cooperation outperforms single-user cooperation in Cop-PD-NOMA communication.

The devil in the tail: Cluster consolidation plus cluster adaptive balancing loss for unsupervised person re-identification

Unsupervised person re-identification (Re-ID) is to retrieve pedestrians from different camera views without supervision information. State-of-the-art methods are usually built upon training a convolution neural network with pseudo labels generated by clustering. Unfortunately, the pseudo labels are highly unbalanced and heavily noisy, carrying ineffective or even erroneous supervision information. To address these deficiencies, we present an effective clustering and reorganization approach, called Cluster Consolidation, which aims to separate a small proportion of unreliable data points from each cluster. This approach benefits to improve the quality of the pseudo labels, but also yields more tiny clusters. Thus, we further propose a Cluster Adaptive Balancing (CAB) loss to effectively train the network with the imbalance pseudo labels, where our CAB loss is able to automatically balance the importance of each cluster. We conduct extensive experiments on widely used person Re-ID benchmark datasets and demonstrate the effectiveness of our proposals.

CBR: An Effective Clustering Approach for Time Series Events

CBR: An Effective Clustering Approach for Time Series Events

Energy-Efficient Routing Protocol for Next-Generation Application in the Internet of Things and Wireless Sensor Networks

Among the key challenges with wireless sensor networks (WSNs) is that most sensor nodes are fueled by energy-constrained batteries, which has a significant impact on the network’s efficiency, reliability, and durability. As a result, many clustering approaches have been developed to enhance the energy efficiency of WSNs. Meanwhile, fifth-generation (5G) transmissions necessitate the usage of multiple-input multiple-output (MIMO) multiple antennas in numerous Internet of Things (IoT) applications to furnish increased capacity in a multipath spectrum environment. Instead of a single senor that can facilitate better load balancing utilization, we believe to balance the energy utilization per unit area. The devices in IoT are submerged with various transmission interfaces known as MIMO in 5G networks. With MIMO being more commonly accessible on IoT devices, an effective clustering approach for rapidly evolving IoT systems is both lacking and urgently needed to support a variety of user scenarios. In this paper, we proposed the intelligent MIMO-based 5G balanced energy-efficient protocol which focuses to achieve Quality of Experience (QoE) for transmitting in clusters for IoT networks. The proposed protocol enhances the utilization of energy and lifetime of the network in which it shows 30% less energy utilized in comparison to the existing protocols.

REFIT: Robustness Enhancement Against Cascading Failure in IoT Networks

There has been tremendous growth in the Internet of Things (IoT) technologies, and many new applications have emerged. However, cascading failure as one of the major issues in such constrained networks have been neglected. In this paper, we apply an effective clustering approach dubbed as REFIT to enhance network topology robustness via nodes’ residual energy. The REFIT protocol divides the network processes into two stages, (i) set-up state and (ii) steady state. The Cluster Head (CH) selection method determines the supreme set of CHs that balances load distribution. The routing method is developed with the modified Particle Swarm Optimization (PSO) algorithm and the objective function to find the supreme set of Relay Nodes (RNs). These complete methods are combined into a set-up state to construct an optimal routing tree that links these CHs to the sink via RNs. In steady state, we model the routing tree to Conditional Directed Acyclic Graph (C-DAG) infrastructure that leads to shortcut routes. Simulation results on MATLAB Simulink have demonstrated that compared with the state-of-the-art works, REFIT can significantly promote network robustness against cascading failure.

Probabilistic density-based estimation of the number of clusters using the DBSCAN-martingale process

Density-based clustering is an effective clustering approach that groups together dense patterns in low- and high-dimensional vectors, especially when the number of clusters is unknown. Such vectors are obtained for example when computer scientists represent unstructured data and then groups them into clusters in an unsupervised way. Another facet of clustering similar artifacts is the detection of densely connected nodes in network structures, where communities of nodes are formulated and need to be identified. To that end, we propose a new DBSCAN algorithm for estimating the number of clusters by optimizing a probabilistic process, namely DBSCAN-Martingale, which involves randomness in the selection of density parameter. We minimize the number of iterations required to extract all clusters by the DBSCAN-Martingale process, by providing an analytic formula. Experiments on spatial, textual and visual clustering show that the proposed analytic formula provides a suitable indicator for the optimal number of required iterations to extract all clusters.

Bootstrap clustering approaches for organization of data: Application in improving grade separability in cervical neoplasia

Abstract This study introduces a novel technique for the self-organization of large datasets, without prior knowledge on the statistical distribution of data. The particular application of interest concerns the self-organization of diffuse reflectance time-lapse curves, expressing the biomarker uptake-wash out kinetics in cervical epithelium. Dynamic spectral imaging generates one curve per pixel resulting to about 700.000 curves per person examined. It comprises an established technology for the non-invasive diagnosis of precancerous lesions, since various curve profiles represent distinct neoplasia grades. The methodology developed in this study constitutes an effective automatic clustering approach improving the discrimination ability between a number of precancerous and non-precancerous abnormalities. The automatic clustering of such a large number of curves is expected to facilitate early diagnosis and prognosis, as well as to assist the management of cervical neoplasia. The effectiveness of the proposed approach stems from the incorporation of data bootstrapping within the clustering approach and the adoption of appropriate distance metrics for assessing class coherence. Each bootstrap step derives a set of cluster centroids, which are then regrouped into active class centers based on a meta-data clustering step. The proposed methodology searches for hidden characteristics within the processed dataset and reveals additional data structures or subclasses that can be utilized for identifying irregular groups, which are of particular importance in disease modeling and management. More specifically, a hidden class was revealed in cervical neoplasia with significant confidence indicated by the metrics of Silhouette, Calonski Harabasz and Dunn´s indices, standard deviation of minimum distance metrics. The results of this study show that appropriate bootstrap extensions of simple clustering schemes can effectively organize large time-series data, giving rise to exploratory approaches for subclass identification that facilitate accurate and early disease diagnosis.

ESCAPE: Effective Scalable Clustering Approach for Parallel Execution of Continuous Position-Based Queries in Position Monitoring Applications

Massive data sets of continuous position-based queries (CPQs) in position monitoring applications offer a challenge of time ingestion while forwarding the position-based data within wireless search space area. As a result of recurrent modifications in network topology due to the mobility of users, processing of large CPQ data sets and roaming CPQs is one of the challenges in position monitoring. Therefore, the parallel algorithm is proposed in this paper for clustering and parallel processing of roaming CPQs which will recognize solid clusters in the wireless search space area. We present an algorithm which proposes the use of search space areas for clustering and introduce a parallel framework for parallel processing of CPQ data sets. The wireless search space area from wireless networks are used for scalable and effective calculation of clusters and dimensions in wireless networks. Present continuous query processing techniques cannot competently process roaming CPQs in the wireless search space area. The proposed algorithm is demonstrated to have practically best speedups in processing roaming CPQs. Results indicate that the proposed work determine improved ability of CPQs over existing mechanisms and attain small query latency and high precision in position monitoring applications.

Towards parameter-independent data clustering and image segmentation

While there are a large amount of clustering algorithms proposed in the literature, the clustering results of existing algorithms usually depend on user-specified parameters heavily, and it is usually difficult to determine the optimal parameters. With the pairwise data similarity matrix as the input, dominant sets clustering has been shown to be an effective data clustering and image segmentation approach, partly due to its ability to find out the underlying data structure and determine the number of clusters automatically. However, we find that the original dominant sets algorithm is sensitive to the similarity measures used in building the similarity matrix. This means that parameter tuning is required to generate satisfactory clustering results, and dominant sets clustering results are also parameter dependent. In order to remove the dependence on the user-specified parameter, we study how the similarity measures influence the dominant sets clustering results. As a result, we propose to transform similarity matrices by histogram equalization before clustering. While this transformation is shown to remove the sensitiveness to similarity measures effectively, it also results in over-segmentation. Therefore in the next step we present a cluster extension method to overcome the over-segmentation effect and generate more reasonable clustering results. We test the enhanced clustering algorithm in both data clustering and image segmentation experiments, and comparisons with the state-of-the-art algorithms validate the effectiveness of our algorithm.

An optimal record retrieval technique in text mining using GSO-based prefix span algorithm and improved K-means

The data mining, nowadays, has surfaced as an investigative procedure devoted to the exploration of the data in the hunt for the reliable patterns or methodical associations between the variables. In the current investigation, an earnest effort made to employ an effective hybrid clustering approach to cluster the record and regain the record in accordance with the pattern mining. The novel technique consists of two vital steps such as the training and testing stages. In the training stage, the closed itemsets of each record are extorted by means of the support values, paving the way for the incredible decrease in the error making items. In the testing stage, the records having identical or approximately identical weights are clustered by means of the hybrid K-means-GSO clustering algorithm. Consequently, records at the apex of rank list are regained from the testing stage. The epoch-making technique is performed in the Java platform.

An Effective Clustering Approach with Data Aggregation Using Multiple Mobile Sinks for Heterogeneous WSN

Wireless Sensor Networks (WSNs) mostly uses static sink to collect data from the sensor nodes randomly deployed in the sensor region. In the static sink based approach, the data packets are flooded across the network to reach the mobile base station in multi-hop communication. Due to this, the static sink is inefficient in energy utilization. Recently, mobile sink are used for data gathering, has less energy utilization which in turn increases the network lifetime. Thus, the sink mobility has difficulties in finding the routing path for the data packets. This paper proposes an effective clustering approach with data aggregation using multiple mobile sinks for heterogeneous WSN. The proposed algorithm achieves network lifetime increases with limited energy utilization.

Clustering and Classification of Documents based on Meta Information using COATES and COLT Algorithms

The side information means the meta information of the documents can be used for the purpose of data mining applications like clustering, classification etc. Huge amount of meta-information is available along with the text documents in many text mining applications. Such meta-information is of different kinds, likes links in the document, user-access behavior from web logs etc. which can be useful for data mining. Tremendous amount of information can be found in this unstructured attributes for clustering purposes. Therefore, this system used an approach which carefully ascertains the coherence of the clustering characteristics of the meta information with that of the text content. For improving the quality of the clustering both the text data and meta information is helpful. In this system, the design of an algorithm which combines classical partitioning algorithms with probabilistic models in order to create an effective clustering approach using meta information present in document was performed. Then it shows how to extend the clustering approach to the classification problem. COATES and COLT algorithm for clustering and classification of text data along with the meta information are used and it shows the advantages of using such an approach. General Terms Data mining, information retrieval, text mining KeywordsClassification, clustering, data mining, meta information, text mining

On the Use of Side Information for Mining Text Data

In many text mining applications, side-information is available along with the text documents. Such side-information may be of different kinds, such as document provenance information, the links in the document, user-access behavior from web logs, or other non-textual attributes which are embedded into the text document. Such attributes may contain a tremendous amount of information for clustering purposes. However, the relative importance of this side-information may be difficult to estimate, especially when some of the information is noisy. In such cases, it can be risky to incorporate side-information into the mining process, because it can either improve the quality of the representation for the mining process, or can add noise to the process. Therefore, we need a principled way to perform the mining process, so as to maximize the advantages from using this side information. In this paper, we design an algorithm which combines classical partitioning algorithms with probabilistic models in order to create an effective clustering approach. We then show how to extend the approach to the classification problem. We present experimental results on a number of real data sets in order to illustrate the advantages of using such an approach.

An Efficient Incremental Clustering based Summarization Technique for Web Page Classification

Currently the World Wide Web is the largest source of information. There are numerous self-acting classification advances that have been suggested. In this suggested work an effective incremental clustering approach to evolve a better incremental clustering based summarization method for world wide World Wide Web classification which can facilitate to better coordinate the accessible data on WWW. The incremental clustering founded summarization technique permits dynamic tracking of the ever expanding allotment of information, put on the World Wide Web every day. This is the useful procedure for dynamic contents. In this the clustering of web document is a key method for finding out a more contextual and noiseless knowledge for world wide world wide web utilizes. C4.5 is one of the most classic classification algorithms on data excavation, when it is utilized in mass computed results, the effectiveness is very low. In this paper, the direct of C4.5 is advanced by the use of L’Hospital direct, which simplifies the assessment process and advances the effectiveness of decision making algorithm. The aim of this work to apply the algorithms in a very time and space effective kind and throughput and answer time for the application will be encouraged as the presentation measures. The aspires is to implement these algorithms and graphically compared the complexities and efficiencies of the classification algorithms.

A fuzzy minimax clustering model and its applications

Fuzzy clustering is an effective clustering approach which associates a data point with multiple clusters. Standard fuzzy clustering models like fuzzy c-means are based on minimizing the total cluster variation, which is defined as the sum of the distances between the data points and their corresponding cluster centers weighted by the membership degrees. In this paper, we propose a fuzzy minimax clustering model by minimizing the maximum value of the set of weighted cluster variations in such a way that they satisfy a prior distribution. We derive a necessary condition for the extremum point of the fuzzy minimax clustering model, and then design an iterative algorithm for solving the extremum point. Several numerical examples on comparing fuzzy c-means and fuzzy minimax clustering models are given, which demonstrate that the prior distribution improves the quality of the clustering results significantly.