Difference between revisions of "Orange: Louvain Clustering"

From OnnoWiki
Jump to navigation Jump to search
Line 3: Line 3:
 
roups items using the Louvain clustering algorithm.
 
roups items using the Louvain clustering algorithm.
  
Inputs
+
==Input==
  
    Data: input dataset
+
Data: input dataset
  
Outputs
+
==Output==
  
    Data: dataset with cluster index as a class attribute
+
Data: dataset with cluster index as a class attribute
    Graph (with the Network addon): the weighted k-nearest neighbor graph
+
Graph (with the Network addon): the weighted k-nearest neighbor graph
  
 
The widget first converts the input data into a k-nearest neighbor graph. To preserve the notions of distance, the Jaccard index for the number of shared neighbors is used to weight the edges. Finally, a modularity optimization community detection algorithm is applied to the graph to retrieve clusters of highly interconnected nodes. The widget outputs a new dataset in which the cluster index is used as a meta attribute.
 
The widget first converts the input data into a k-nearest neighbor graph. To preserve the notions of distance, the Jaccard index for the number of shared neighbors is used to weight the edges. Finally, a modularity optimization community detection algorithm is applied to the graph to retrieve clusters of highly interconnected nodes. The widget outputs a new dataset in which the cluster index is used as a meta attribute.
Line 16: Line 16:
 
[[File:Louvain-stamped.png|center|200px|thumb]]
 
[[File:Louvain-stamped.png|center|200px|thumb]]
  
    PCA processing is typically applied to the original data to remove noise.
+
* PCA processing is typically applied to the original data to remove noise.
    The distance metric is used for finding specified number of nearest neighbors.
+
* The distance metric is used for finding specified number of nearest neighbors.
    The number of nearest neighbors to use to form the KNN graph.
+
* The number of nearest neighbors to use to form the KNN graph.
    Resolution is a parameter for the Louvain community detection algorithm that affects the size of the recovered clusters. Smaller resolutions recover smaller, and therefore a larger number of clusters, and conversely, larger values recover clusters containing more data points.
+
* Resolution is a parameter for the Louvain community detection algorithm that affects the size of the recovered clusters. Smaller resolutions recover smaller, and therefore a larger number of clusters, and conversely, larger values recover clusters containing more data points.
    When Apply Automatically is ticked, the widget will automatically communicate all changes. Alternatively, click Apply.
+
* When Apply Automatically is ticked, the widget will automatically communicate all changes. Alternatively, click Apply.
  
 
==Contoh==
 
==Contoh==

Revision as of 08:40, 29 January 2020

Sumber: https://docs.biolab.si//3/visual-programming/widgets/unsupervised/louvainclustering.html

roups items using the Louvain clustering algorithm.

Input

Data: input dataset

Output

Data: dataset with cluster index as a class attribute
Graph (with the Network addon): the weighted k-nearest neighbor graph

The widget first converts the input data into a k-nearest neighbor graph. To preserve the notions of distance, the Jaccard index for the number of shared neighbors is used to weight the edges. Finally, a modularity optimization community detection algorithm is applied to the graph to retrieve clusters of highly interconnected nodes. The widget outputs a new dataset in which the cluster index is used as a meta attribute.

Louvain-stamped.png
  • PCA processing is typically applied to the original data to remove noise.
  • The distance metric is used for finding specified number of nearest neighbors.
  • The number of nearest neighbors to use to form the KNN graph.
  • Resolution is a parameter for the Louvain community detection algorithm that affects the size of the recovered clusters. Smaller resolutions recover smaller, and therefore a larger number of clusters, and conversely, larger values recover clusters containing more data points.
  • When Apply Automatically is ticked, the widget will automatically communicate all changes. Alternatively, click Apply.

Contoh

Louvain Clustering converts the dataset into a graph, where it finds highly interconnected nodes. We can visualize the graph itself using the Network Explorer from the Network addon.

Louvain-Example.png

Referensi

Blondel, Vincent D., et al. “Fast unfolding of communities in large networks.” Journal of statistical mechanics: theory and experiment 2008.10 (2008): P10008.

Lambiotte, Renaud, J-C. Delvenne, and Mauricio Barahona. “Laplacian dynamics and multiscale modular structure in networks.” arXiv preprint, arXiv:0812.1770 (2008).


Referensi

Pranala Menarik