With the development of abnormal behavior analysis technology, measuring the similarity of abnormal behavior has become a core part of abnormal behavior detection. However, there are general problems of central select...
详细信息
ISBN:
(纸本)9783030005634;9783030005627
With the development of abnormal behavior analysis technology, measuring the similarity of abnormal behavior has become a core part of abnormal behavior detection. However, there are general problems of central selection distortion and slow iterative convergence with existing clusteringbased analysis algorithms. Therefore, this paper proposes an improved clustering-based abnormal behavior analysis algorithm by using K-means. Firstly, an abnormal behavior set is constructed for each user from his or her behavioral data. A weight calculation method for abnormal behaviors and an eigenvalue extraction method for abnormal behavior sets are proposed by using all the behavior sets. Secondly, an improved algorithm is developed, in which we calculate the tightness of all data points and select the initial cluster centers from the data points with high density and low density to improve the clustering effect based on the K-means clustering algorithm. Finally, clustering result of the abnormal behavior is got with the input of the eigenvalues of the abnormal behavior set. The results show that, the proposed algorithm is superior to the traditional clustering algorithm in clustering performance, and can effectively enhance the clustering effect of abnormal behavior.
The past decade has seen a dramatic improvement in the quality of data available at both high(HE,>10 Ge V)and very high(VHE,>100 Ge V)gamma-ray *** to the latest Pass8 data release by Fermi LAT which increases
The past decade has seen a dramatic improvement in the quality of data available at both high(HE,>10 Ge V)and very high(VHE,>100 Ge V)gamma-ray *** to the latest Pass8 data release by Fermi LAT which increases
Smart grid has great application prospects in the power industry. How to extend network life has become a research hotspot in recent years. This paper improves the wireless sensor network LEACH algorithm by analyzing ...
详细信息
Smart grid has great application prospects in the power industry. How to extend network life has become a research hotspot in recent years. This paper improves the wireless sensor network LEACH algorithm by analyzing the business requirements of the Smart grid. First, the remaining energy and node density are taken into account in the cluster head election, to ensure the rationality of the cluster head selection. Then, a non-uniform clustering mechanism is adopted to construct the cluster radius of different sizes when clustering. The multi-hop mode is used to transmit data to the base station, and the distance between the cluster head nodes and the remaining energy are used as the selection basis of the next-hop relay node to achieve the purpose of equalizing the energy load. Simulation results show that compared with LEACH and LEACH-DC algorithms, the proposed algorithm can effectively balance node energy consumption and extend network life cycle.
There is great need to estimate the number of sources before the blind source separation technique *** on the clustering algorithm in machine learning,this study constructs the inter-frequency distance according to th...
详细信息
There is great need to estimate the number of sources before the blind source separation technique *** on the clustering algorithm in machine learning,this study constructs the inter-frequency distance according to the magnitude of each frequency in the signal spectrum curve,uses clustering algorithms such as Kmeans,DBSCAN and Gaussian mixture to cluster the number of frequency bands of the observed signals,then analysis the comparative performance.
Aiming at the problem of video key frame extraction, a density peak clustering algorithm is proposed, which uses the HSV histogram to transform high-dimensional abstract video image data into quantifiable low-dimensio...
详细信息
Aiming at the problem of video key frame extraction, a density peak clustering algorithm is proposed, which uses the HSV histogram to transform high-dimensional abstract video image data into quantifiable low-dimensional data, and reduces the computational complexity while capturing image features. On this basis, the density peak clustering algorithm is used to cluster these low-dimensional data and find the cluster centers. Combining the clustering results, the final key frames are obtained. A large number of key frame extraction experiments for different types of videos show that the algorithm can extract different number of key frames by combining video content, overcome the shortcoming of traditional key frame extraction algorithm which can only extract a fixed number of key frames, and the extracted key frames can represent the main content of video accurately.
Since spectrum environment is increasingly complex and more devices accessed communication networks, traditional spectrum sensing algorithms have not adapted to an extreme volume of spectrum data. In this paper, a nov...
详细信息
ISBN:
(纸本)9781450365765
Since spectrum environment is increasingly complex and more devices accessed communication networks, traditional spectrum sensing algorithms have not adapted to an extreme volume of spectrum data. In this paper, a novel deep cooperative spectrum sensing scheme is proposed, which combined principal component analysis (PCA) with clustering algorithm. First, a multi-dimension feature matrix, consisting of energy vectors generated in fusion center, is reduced to a lower-dimension matrix according to the PCA algorithm. Subsequently, a clustering algorithm with K-means++ method is utilized to train the classifier by the lower-dimensional matrix. The simulation results show that the proposed scheme has shorter training duration about 64% of no PCA processing when the primary user power is 400 mW, and ensures spectrum sensing accuracy of secondary users. More importantly, the proposed scheme, compared with the others' cooperative spectrum sensing schemes, can significantly reduce the required hardware memory.
The ability to process data streams has become one of the challenges of the current intrusion detection systems. A data stream clustering algorithm based on bucket density is proposed for this situation which is able ...
详细信息
ISBN:
(纸本)9789811076053;9789811076046
The ability to process data streams has become one of the challenges of the current intrusion detection systems. A data stream clustering algorithm based on bucket density is proposed for this situation which is able to identify clusters in any shapes and the speed of online layer is fast. Feedback principle is used to solve the problem that some of the edge of the bucket is lost and users does not need to specify the number of clusters. An intrusion detection system is constructed with the improved algorithm. The experiment shows that the algorithm proposed has fast speed for clustering. The system based on the algorithm has a better capability of detection.
A low-complexity clustering algorithm can achieve modulation recognition and resistant nonlinearity distortion of RoF System is proposed. This technique can classify different modulations accurately for any order and ...
详细信息
ISBN:
(纸本)9781538691458
A low-complexity clustering algorithm can achieve modulation recognition and resistant nonlinearity distortion of RoF System is proposed. This technique can classify different modulations accurately for any order and can improve the BER performance.
Traditional fingerprint orientation clustering algorithms often use k means clustering algorithm, but as a result of fingerprint and objective factors of volatile characteristics over time, k-means cannot adapt to cha...
详细信息
ISBN:
(纸本)9783319745213;9783319745206
Traditional fingerprint orientation clustering algorithms often use k means clustering algorithm, but as a result of fingerprint and objective factors of volatile characteristics over time, k-means cannot adapt to change at any time in fingerprint, and cannot be generated adaptive clustering cluster number, cause the matching accuracy is not high. This paper adopts a based on support vector machine (SVM) and DBSCAN clustering algorithm, can generate continuously adapt to changing the optimal hyperplane fingerprint model, solved the fingerprint fluctuating lead to the problem of matching result is bad, and can be automatically generated in the process of matching classification number of clusters, based on statistical density characteristics of DBSCAN selection matching probability model, to improve the positioning of the matching accuracy, reduced the amount of time matching positioning, positioning accuracy can be up to 2.04 m in the range of 57%, relative k-means 6.1 m increased by 52.3%, improve the positioning accuracy.
This paper builds upon our previous paper that has introduced a new hierarchical clustering algorithm. In this paper we attempted to solve algorithmic defects and use more validation measures to show the strength of o...
详细信息
ISBN:
(纸本)9781538650028
This paper builds upon our previous paper that has introduced a new hierarchical clustering algorithm. In this paper we attempted to solve algorithmic defects and use more validation measures to show the strength of our proposed algorithm. The main purpose of this clustering algorithm is to provide a better clustering quality and higher accuracy utilizing intersection points. To validate our clustering algorithm, we have performed several experiments with benchmark datasets. Besides our proposed algorithm, five well-known agglomerative clustering algorithms are also used. Purity as an external criterion is used to evaluate the performance of clustering algorithms. Compactness of each cluster derived by clustering algorithms is also calculated to evaluate the validity of clustering algorithms. Eventually, the results of experiments show that in most cases the error rate of our proposed algorithm is lower than other clustering algorithms which are used in this study.
暂无评论