With(1) the rapid development of information society, data streams have become the main data model in many fields. In order to dig out the useful information contained in data, these data stream clustering algorithms ...
详细信息
ISBN:
(纸本)9781450365123
With(1) the rapid development of information society, data streams have become the main data model in many fields. In order to dig out the useful information contained in data, these data stream clustering algorithms are particularly important. There are two key issues in the process of handing data stream with data stream clustering algorithm: On the one hand, it is how to judge outliers;on the other hand, it is how to eliminate outdated data in time. Aiming at these two problems, this paper proposes a DCluStream algorithm. The algorithm mainly is designed a set of buffer processing mechanism to deal with abnormal data in order to correctly judge whether these abnormal data are outliers. In addition, the DCluStream algorithm is added the decay time window in the stage of the online micro-clustering, and each data is assigned weight value. Through observing real-time weight of micro cluster for each micro cluster, the algorithm eliminates these overdue micro clusters in time and better deals with recent data in order to realize the accurate clustering. Finally, the DClustream algorithm uses KDD CUP99 data set for simulation experiments. These experimental results show that the new algorithm improves the clustering quality and reduces the clustering processing time, as well as it cuts down memory occupancy.
The initial clustering center of the traditional K-means algorithm is randomly generated from the data set which leads it easily gets the local optimal solution rather than the global optimal solution. In this paper, ...
详细信息
ISBN:
(纸本)9781538643624
The initial clustering center of the traditional K-means algorithm is randomly generated from the data set which leads it easily gets the local optimal solution rather than the global optimal solution. In this paper, a high-dimensional quantum genetic clustering method is proposed in which each quantum bit is decomposed into a plurality of parallel genes according to a high-dimensional quantum encoding scheme and it effectively expands the search space and enhances the efficiency of parallel search. A quantum updating strategy comes out by combining the high-dimensional coding scheme and the dynamically adjustment rotation angle mechanism to renew the individual. Quantum mutation is implemented by quantum non-gate mutation strategy to enhance the global searching ability of the algorithm. Experiments show that the algorithm is not only better in clustering accuracy, but also faster in convergence.
In order to improve the spectrum sensing performance, we propose a cooperative spectrum sensing method based on a feature and clustering algorithm in the case of a small number of secondary users participating in coop...
详细信息
ISBN:
(纸本)9781728113128
In order to improve the spectrum sensing performance, we propose a cooperative spectrum sensing method based on a feature and clustering algorithm in the case of a small number of secondary users participating in cooperative spectrum sensing. This method introduces order decomposition and recombination and interval decomposition and recombination based on stochastic matrix, which can increases the secondary users logically. Firstly, the signal matrix collected by the secondary users is split and recombined, and the corresponding covariance matrix are calculated respectively to obtain the corresponding signal features. Based on these features, we construct them as a feature vector. Further, we will use the clustering algorithm to train and perform spectrum sensing based on the trained classifier. In the experimental and results analysis section, the method described in this paper was simulated and the experimental results were further analyzed.
To explore the Internet of things logistics system application, an Internet of things big data clustering analysis algorithm based on K-mans was discussed. First of all, according to the complex event relation and pro...
详细信息
To explore the Internet of things logistics system application, an Internet of things big data clustering analysis algorithm based on K-mans was discussed. First of all, according to the complex event relation and processing technology, the big data processing of Internet of things was transformed into the extraction and analysis of complex relational schema, so as to provide support for simplifying the processing complexity of big data in Internet of things (IOT). The traditional K-means algorithm was optimized and improved to make it fit the demand of big data RFID data network. Based on Hadoop cloud cluster platform, a K-means cluster analysis was achieved. In addition, based on the traditional clustering algorithm, a center point selection technology suitable for RFID IOT data clustering was selected. The results showed that the clustering efficiency was improved to some extent. As a result, an RFID Internet of things clustering analysis prototype system is designed and realized, which further tests the feasibility.
As a special form of Mobile Ad hoc NETwork (MANET), vehicles in the network can connect to other vehicles on the road and the Internet, and can provide stable and high-speed wireless data access services for vehicles ...
详细信息
ISBN:
(纸本)9781538676356
As a special form of Mobile Ad hoc NETwork (MANET), vehicles in the network can connect to other vehicles on the road and the Internet, and can provide stable and high-speed wireless data access services for vehicles with high velocity. VANET has become an effective technology and important means to guarantee vehicle safety, provide intelligent traffic management of high-speed data communication and vehicle entertainment. However, the typical characteristics of VANET, including rapidly changing network topology, highly dynamic channel condition and node competition in channel accessing, etc., raise difficulties and challenges on data transmission in VANET. To stress the problem, a cluster algorithm based on vehicle mobility for VANET is proposed, in the proposed algorithm, the object function of cluster head selection is formulated based on the vehicle mobility including position, velocity and packet forwarding capability, the process of the clustering algorithm is also presented. Simulations demonstrate that comparing to previously proposed algorithms, the proposed algorithm offers better performance in terms of packet delivery rate, average transmission delay and total throughput.
Considering the problem of autonomous course-tracking of Unmanned Surface Vessel(USV) with uncertain systems,this paper put forwards a control method based on the clustering algorithm. The designed controller is compo...
详细信息
Considering the problem of autonomous course-tracking of Unmanned Surface Vessel(USV) with uncertain systems,this paper put forwards a control method based on the clustering algorithm. The designed controller is composed of a clustering algorithm controller and a PID controller. Because of the uncertainty of the system, the clustering algorithm analysis controller online learned and classified the expected course to control the rudder, the PID controller adjusts the error. This method solves the problems of low control accuracy and poor control performance of robustness brought by the single system controller. By applying the proposed control method to the control design of USV autonomous course-tracking, the results of the simulation experiment show the effectiveness of the control method.
A simple mechanism to prolong the life cycle of the network by balancing nodes' energy consumption is to rotate the active dominating set (DS) through a set of legitimate DSs. This paper proposes a novel adaptive ...
详细信息
A simple mechanism to prolong the life cycle of the network by balancing nodes' energy consumption is to rotate the active dominating set (DS) through a set of legitimate DSs. This paper proposes a novel adaptive clustering algorithm named HREF (Highest Remaining Energy First). In the HREF algorithm, cluster formation is performed cyclically and each node can declare itself as a cluster head autonomously if it has the largest residual energy among all its adjacent nodes. The performance effectiveness of the HREF algorithm is investigated and compared to the D-WCDS (Disjoint Weakly Connected Dominating Set) algorithm. In this paper, we assume the network topology is fixed and does not require sensor mobility. This allows us to focus on the impact of clustering algorithms on communication between network nodes rather than with the base station. Simulation results show that in the D-WCDS algorithm energy depletion is more severe and the variance of the node residual energy is also much larger than that in the HREF algorithm. That is, nodes' energy consumption in the HREF algorithm is in general more evenly distributed among all network nodes. This may be regarded as the main advantage of the HREF adaptive clustering algorithm.
In processing of large scene synthetic aperture radar (SAR) images, the first step is to split them into tiles in order to reduce the load of computer's computation effort and memory, which is crucial in the follo...
详细信息
ISBN:
(纸本)9781538691540
In processing of large scene synthetic aperture radar (SAR) images, the first step is to split them into tiles in order to reduce the load of computer's computation effort and memory, which is crucial in the follow-up procedures of building radar footprints detection or reconstruction. Compared to the traditional cockamamie gridding method, we propose an automatic sub-images extraction approach based on the density and distance-based (DD) clustering algorithm and the connected-component labeling (CCL) algorithm of the graph theory which can avoid a mass of unnecessary least error finding work. According to our method, the original image can be split without introducing excessive subjective operation, which can remain the primary information of buildings' images efficiently. Meanwhile, automatic area searching reduces manpower effectively.
The biological signals collected by the multi-electrode array are contaminated by heavy noise signals. How to quickly classify the original action potential from the measured noisy signals accurately is the basis of r...
详细信息
ISBN:
(纸本)9781538635247
The biological signals collected by the multi-electrode array are contaminated by heavy noise signals. How to quickly classify the original action potential from the measured noisy signals accurately is the basis of researches in the field of neuroscience. In this paper, we analyze the characteristics and shortcomings of Wave-clus sorting algorithm, and present a novel sorting algorithm to solve the classification problem of neuronal action potentials. Besides, a novel method of FPGA implementation is introduced for the clustering algorithm. The method and simulation results of estimating the waveform function model are given. The prediction electrode data and the algorithm of the residual function module are set up and the error analysis is carried out. The simulation results show that the proposed method has a satisfied performance.
As one of the most important techniques in data mining, cluster analysis has attracted more and more attentions in this big data era. Most clustering algorithms have encountered with challenges including cluster cente...
详细信息
As one of the most important techniques in data mining, cluster analysis has attracted more and more attentions in this big data era. Most clustering algorithms have encountered with challenges including cluster centers determination difficulty, low clustering accuracy, uneven clustering efficiency of different data sets and sensible parameter dependence. Aiming at clustering center determination difficulty and parameter dependence, a novel cluster center fast determination clustering algorithm was proposed in this paper. It is supposed that clustering centers are those data points with higher density and larger distance from other data points of higher density. Normal distribution curves are designed to fit the density distribution curve of density distance product. And the singular points outside the confidence interval by setting the confidence interval are proved to be clustering centers by theory analysis and simulations. Finally, according to these clustering centers, a time scan clustering is designed for the rest of the points by density to complete the clustering. Density radius is a sensible parameter in calculating density for each data point, mountain climbing algorithm is thus used to realize self-adaptive density radius. Abundant typical benchmark data sets are testified to evaluate the performance of the brought up algorithms compared with other clustering algorithms in both aspects of clustering quality and time complexity. (C) 2017 Published by Elsevier B.V.
暂无评论