Search

Data analysis becomes more important since rapid technology developments. In data analysis, data clustering is one of the very useful approaches. It can reveal important information hiding inside the dataset by organizing the instances based on their similarity. The objectives of data clustering are maximizing dissimilarity between clusters and minimizing dissimilarity within clusters. In order to construct a good clustering results, many clustering algorithms have been proposed, including the metaheuristic-based clustering algorithms. Recently, a new metaheuristic algorithm named gradient evolution has been proposed. This algorithm shows a good performance on solving the optimization problems. Therefore, this paper employs this GE algorithm for solving the clustering problem. In order to obtain a better clustering result, this paper considers multi-objective clustering instead of single-objective clustering. In this paper, the original GE algorithm is improved so then it is suitable for the multi-objective problem. The proposed modification includes the procedure for vector updating and jumping which involves Pareto rank assignment. In addition, it also employs K-means algorithm to provide the final clustering result. The proposed algorithm is verified using some benchmark datasets. It is also compared with some other multi-objective metaheuristic-based clustering algorithms. The experimental results show that the proposed algorithm can obtain better results than other metaheuristic-based algorithms. [ABSTRACT FROM AUTHOR]

Clustering is an important data mining technique described as unsupervised learning. Till date, many single-objective clustering algorithms have been developed on the basis of swarm intelligence and evolutionary techniques. It is noticed that these clustering algorithms provide better solutions for clustering problems, but sometimes, these solutions seem to be biased and also not appropriate for the problem with geometrical shapes datasets. In turn, performance of the clustering algorithms can be degraded. One of the possible solutions is to adopt multi-objective approach instead of single objective. In multi-objective approach, more than one objective functions can be considered for solving the clustering problems and these functions are conflicted in nature. Further, in multi-objective approach, Pareto-optimal solutions can be generated for improving the clustering performance. Hence, this paper presents a multi-objective clustering algorithm based on vibrating particle system (VPS) for effective cluster analysis, called MOVPS. This work considers intra-cluster variance and connectedness as objective functions, and VPS algorithm is used for optimizing the aforementioned objectives to obtain good clustering results. The performance of MOVPS algorithm is tested over a set of benchmark datasets and validated by comparing clustering results with various multi-objective and single-objective clustering algorithms from the literature. The simulation results illustrate the effectiveness of the MOVPS algorithm based on F-measure, coverage, distribution, convergence, non-dominating vector generation and intra-cluster distance measures. The simulation results showed that the proposed MOVPS algorithm enhances the clustering results significantly in comparison with existing multi-objective and single-objective clustering algorithms. [ABSTRACT FROM AUTHOR]

This paper proposes an Automatic Clustering algorithm for Interval data using the Genetic algorithm (ACIG). In this algorithm, the overlapped distance between intervals is applied to determining the suitable number of clusters. Moreover, to optimize in clustering, we modify the Davies & Bouldin index, and to improve the crossover, mutation, and selection operators of the original genetic algorithm. The convergence of ACIG is theoretically proved and illustrated by the numerical examples. ACIG can be implemented effectively by the established Matlab procedure. Through the experiments on data sets with different characteristics, the proposed algorithm has shown the outstanding advantages in comparison to the existing ones. Recognizing the images by the proposed algorithm gives the potential in real applications of this research. [ABSTRACT FROM AUTHOR]

Fuzzy C-means (FCM) is a classical algorithm of cluster analysis which has been applied to many fields including artificial intelligence, pattern recognition, data aggregation and their applications in software engineering, image processing, IoT, etc. However, it is sensitive to the initial value selection and prone to get local extremum. The classification effect is also unsatisfactory which limits its applications severely. Therefore, this paper introduces the artificial-fish-swarm algorithm (AFSA) which has strong global search ability and adds an adaptive mechanism to make it adaptively adjust the scope of visual value, improves its local and global optimization ability, and reduces the number of algorithm iterations. Then it is applied to the improved FCM which is based on the Mahalanobis distance, named as adaptive AFSA-inspired FCM(AAFSA-FCM). The optimal solution obtained by adaptive AFSA (AAFSA) is used for FCM cluster analysis to solve the problems mentioned above and improve clustering performance. Experiments show that the proposed algorithm has better clustering effect and classification performance with lower computing cost which can be better to apply to every relevant area, such as IoT, network analysis, and abnormal detection. [ABSTRACT FROM AUTHOR]

Intrusion detection systems (IDSs) are devices or software applications that monitor networks or systems for malicious activities and signals alerts/alarms when such activity is discovered. However, an IDS may generate many false alerts which affect its accuracy. In this paper, we develop a cyberattack triage algorithm to detect these alerts (so-called outliers). The proposed algorithm is designed using the clustering, optimization and distance-based approaches. An optimization-based incremental clustering algorithm is proposed to find clusters of different types of cyberattacks. Using a special procedure, a set of clusters is divided into two subsets: normal and stable clusters. Then, outliers are found among stable clusters using an average distance between centroids of normal clusters. The proposed algorithm is evaluated using the well-known IDS data sets—Knowledge Discovery and Data mining Cup 1999 and UNSW-NB15—and compared with some other existing algorithms. Results show that the proposed algorithm has a high detection accuracy and its false negative rate is very low. [ABSTRACT FROM AUTHOR]

Clustering by fast search and find of density peaks (herein called FDPC), as a recently proposed density-based clustering algorithm, has attracted the attention of many researchers since it can recognize arbitrary-shaped clusters. In addition, FDPC needs only one parameter d c and identifies the number of clusters by decision graph. Nevertheless, it is not clear how to find a proper d c for a given data set and such a perfect parameter may not exist in practice for the multi-scale data set. In this paper, we proposed a modified PageRank algorithm to compute the local density for each data point which is more robust than Gaussian kernel and cutoff method. Besides, FDPC yields poor results on the random distribution data sets since there may be several maxima for one cluster. To solve this problem, we proposed an improved minimax similarity method. Comparing our proposed approach with FDPC on some artificial and real-life data sets, the experimental results indicate that our proposed approach outperforms FDPC in terms of accuracy. [ABSTRACT FROM AUTHOR]

Membrane computing is widely used in many areas, however, there are several limitations in its structures and rules. Although many researchers are engaged in the study of P systems, seldom focus on improving membrane structures. The purpose of this paper is to propose a new kind of communication P system on lattice (LTC-P systems). We describe membrane structures on lattice with communication rules. The computational completeness of the new P system is proved by simulation of register machine. The new P system is used in solving clustering problems. It is combined with the thought of density-based, partition-based and hierarchical clustering algorithm. Clustering is implemented by supremum and infimum rules. The result is obtained through output membrane. All the processes are conducted in membranes. Cluster result via a $$20$$ points data set verifies that the proposed new P systems cluster data set accurately and reduce time complexity. Wine data set are also used in testing the influence of parameters. More suitable $$\varepsilon $$ and $${ MinPts}$$ are found to gain less missing data which are seen as noise. Comparative results in various aspects indicate LTC-P system based clustering algorithm consumes less time than traditional algorithms significantly. It also uses less rules and gives more simple membrane structures than conventional cell-like P system. The new P system provides an alternative for traditional membrane computing. [ABSTRACT FROM AUTHOR]