1887
banner image
No data available.
Please log in to see this content.
You have no subscription access to this content.
No metrics data to plot.
The attempt to load metrics for this article has failed.
The attempt to plot a graph for these metrics has failed.
Complex network classification using partially self-avoiding deterministic walks
Rent:
Rent this article for
USD
10.1063/1.4737515
/content/aip/journal/chaos/22/3/10.1063/1.4737515
http://aip.metastore.ingenta.com/content/aip/journal/chaos/22/3/10.1063/1.4737515

Figures

Image of FIG. 1.
FIG. 1.

Regular lattice with random weights assigned to the first and second closed neighbors. An agent leaves a given site and moves according to the rule of not returning to the last μ visited sites. The trajectory is composed of a transient time (orange sites) and an attractor (green sites).

Image of FIG. 2.
FIG. 2.

Example of the first steps of an agent using memory of size μ = 1 and movement rule of going to the closest vertex.

Image of FIG. 3.
FIG. 3.

Example of the first steps of an agent using memory of size μ = 1 and movement rule of going to the furthest vertex.

Image of FIG. 4.
FIG. 4.

Histograms of walk length for different complex network models built using N = 50000 and vertices degree mean .

Image of FIG. 5.
FIG. 5.

Histograms of walk length for number of vertices varying from N = 10000 to 50000 on the small-world model.

Image of FIG. 6.
FIG. 6.

Histograms of walk length for degree mean varying from to 50 on the small-world model.

Image of FIG. 7.
FIG. 7.

PCA projection for 4000 networks obtained by using Erdõs-Rényi, geographical network, small-world and scale-free. Networks were built with N = 1000 and , and the walks were performed with μ = 5. (a) The walker chooses to go to the closest site, and (b) the walker goes to the furthest one.

Image of FIG. 8.
FIG. 8.

PCA projection for complex network models built with N = 1000 and using deterministic walks with different values of memory and din = [min, max]. (a) μ = 0, (b) μ = 1, (c) μ = 2, (d) μ = 3, (e) μ = 4 and (f) μ = 5.

Image of FIG. 9.
FIG. 9.

PCA projection of signature vectors composed by the concatenation of memories 0, 1, 2, 3, 4, and 5. The signatures vectors were extracted from complex networks built with N = 1000 and .

Image of FIG. 10.
FIG. 10.

Classification results for different percentages on the number of vertices.

Image of FIG. 11.
FIG. 11.

Scatter plot matrix for the first three features obtained from 4000 complex networks.

Image of FIG. 12.
FIG. 12.

Dendrogram of the class means using the manova takes a set of grouped data composed by the features extracted using the partially self-avoiding walks characteristics of 1000 complex networks for each class (random network, scale-free, geographical, and small-world).

Image of FIG. 13.
FIG. 13.

Classification results using different values of α in the nonlinear Barabási-Albert model. When α = 1, the nonlinear Barabási-Albert and the Barabási-Albert model are the same.

Tables

Generic image for table
Table I.

Statistics of the histograms for different complex network models.

Generic image for table
Table II.

Statistics of the histograms for different values of N on the small-world model.

Generic image for table
Table III.

Statistics of the histograms for different values of on the small-world model.

Generic image for table
Table IV.

Correct classification rate for with different values of μ and movement rule din.

Generic image for table
Table V.

Correct classification rate for composed by the concatenation of values μ.

Generic image for table
Table VI.

Classification results of different scale-free models.

Generic image for table
Table VII.

Comparison between measurements extracted from the complex networks.

Generic image for table
Table VIII.

Comparison between combination of measurements.

Loading

Article metrics loading...

/content/aip/journal/chaos/22/3/10.1063/1.4737515
2012-09-04
2014-04-24
Loading

Full text loading...

This is a required field
Please enter a valid email address
752b84549af89a08dbdd7fdb8b9568b5 journal.articlezxybnytfddd
Scitation: Complex network classification using partially self-avoiding deterministic walks
http://aip.metastore.ingenta.com/content/aip/journal/chaos/22/3/10.1063/1.4737515
10.1063/1.4737515
SEARCH_EXPAND_ITEM