Distributed clustering method of P2P (peer-to-peer) network based on believable radius of nodes
A technology of distributed clustering and P2P network, which is applied in the field of network) distributed clustering, which can solve the problems of large bandwidth consumption, high network bandwidth usage, and no consideration of data distribution and locality, etc., so as to save bandwidth and improve Application level, the effect of ensuring the clustering effect
Inactive Publication Date: 2011-09-14
JIANGSU UNIV
View PDF3 Cites 4 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0004]In the distributed clustering method in the P2P environment, Eisenhardt et al. proposed a distributed clustering method of Probe / Echo (detection / response) mechanism, which The method achieves a more accurate clustering effect by synchronizing node clustering in the network, but the defect of this method is that the network bandwidth usage rate is high and the bandwidth consumption is very large
Jin et al. proposed the DFEKM method, which completes clustering by calculating the confidence radius of clustering, but this method does not propose a suitable confidence radius calculation method, and uses a fixed threshold parameter for all nodes to determine the size of the confidence radius. Since clustering can only be performed locally on distributed nodes, the fixed radius set by this method does not consider the characteristics of data distribution and locality on nodes, and still requires more iterations to complete clustering
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment
[0087] Take 60,000 data from 3 types of 2-dimensional Gaussian mixture distributions as data objects. The mean values of the Gaussian distributions used are: (0, 0), (6, 2), (8, 8), and the covariance matrix is Two-dimensional identity matrix, set 500 network nodes at the same time, distribute 60,000 data objects to 500 network nodes evenly, and set the k value of clustering to 3.
[0088] Figure 4 It describes the distribution map of the 60,000 data used as data clustering objects. Among them, different colors and shapes represent different Gaussian distribution data, diamonds, stars and triangles represent the mean value of Gaussian distribution data at (0, 0), (6, 2), (8, 8), and the covariance matrix is two Three types of data in the dimensional identity matrix.
[0089] Apply the above data to 500 network nodes, and the results obtained by running the DFEKM method and the method of the present invention 50 times respectively are as follows Figure 5 and Image 6 ...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses a distributed clustering method of a P2P (peer-to-peer) network based on believable radius of nodes, which is characterized in that a distributed K-Means clustering algorithm is adopted by using a Fisher discrimination method to determine believable radius of clusters so as to carry out autonomic learning on data of each P2P node; nodes in the network respectively apply Fisher linear discriminant rate to determine dense and sparse distribution of the same type of data on the nodes, thereby determining the believable radius of the cluster and guiding the next cluster; and according to data distribution of each node in the network, the believable radius are calculated dynamically, thereby determining the believable radius of the clusters of each node and guiding the next cluster and the iterative process of the cluster. By using the distributed clustering method, at the same time of ensuring the clustering effects, the iterations of the clusters in the distributed network can be decreased so as to save bandwidth, and the application level of the network is improved.
Description
technical field [0001] The invention relates to the fields of computer network communication, data mining and distributed clustering, in particular to a P2P network (peer-to-peer network) distributed clustering method. Background technique [0002] Clustering analysis refers to the process of dividing a given set of data objects into multiple categories according to certain rules. Clustering makes the data objects in the same cluster as similar as possible to each other, and the data objects in different clusters Data objects are as distinct from each other as possible. So far, many clustering methods have been proposed, such as K-Means, DBSCAN, Cure, Birch, etc. Among many clustering methods, K-Means is one of the most widely used methods, because compared with other methods, K-Means method has the characteristics of simple method, easy implementation and stable clustering effect. [0003] The analysis and processing of the traditional K-Means clustering method is based o...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More IPC IPC(8): H04L29/08H04L12/58
Inventor沈项军蒋中秋林琳朱倩张科泽
OwnerJIANGSU UNIV