AutomaticProjectionBasedClustering: Automatic Projection-Based Clustering

In FCPS: Fundamental Clustering Problems Suite

Automatic Projection-Based Clustering

Description

Projection-based clustering AutomaticProjectionBasedClustering projects the data (nonlinear) into two dimensions and tries only to preserve relevant neighborhoods prior to clustering. The cluster analysis itself includes the high-dimensional distances in the clustering process. Performs non-interactive projection-based clustering based on non-linear projection methods [Thrun/Ultsch, 2017], [Thrun/Ultsch, 2020a].

Usage

AutomaticProjectionBasedClustering(DataOrDistances,ClusterNo,Type="NerV",

StructureType = TRUE,PlotIt=FALSE,PlotTree=FALSE,PlotMap=FALSE,...)

Arguments

DataOrDistances	Either nonsymmetric [1:n,1:d] numerical matrix of a dataset to be clustered. It consists of n cases of d-dimensional data points. Every case has d attributes, variables or features. or symmetric [1:n,1:n] distance matrix, e.g. as.matrix(dist(Data,method))
ClusterNo	A number k which defines k different clusters to be built by the algorithm.
Type	Type of Projection method, either NerV [Venna et al., 2010] Pswarm [Thrun/Ultsch, 2020b] MDS [Torgerson, 1952] Uwot [McInnes et al., 2018] CCA [Demartines/Herault, 1995] Sammon [Sammon, 1969] t-SNE [Van der Maaten/Hinton, 2008]
StructureType	Either compact (TRUE) or connected (FALSE), see discussion in [Thrun, 2018]
PlotIt	Default: FALSE, if TRUE plots the first three dimensions of the dataset with colored three-dimensional data points defined by the clustering stored in Cls
PlotTree	TRUE: Plots the dendrogram, FALSE: no plot
PlotMap	Plots the topographic map [Thrun et al., 2016].
...	Further arguments to be set for the clustering algorithm, if not set, default arguments are used.

Details

The first idea of using non-PCA projections for clustering was published by [Bock, 1987] as a definition. However, to the knowledge of the author, it was not applied to any data. The coexistence of projection and clustering was introduced in [Thrun/Ultsch, 2017].

Projection-based clustering is based on a nonlinear projection of high-dimensional data into a two-dimensional space [Thrun/Ultsch, 2020b]. Typical projection-methods like t-distributed stochastic neighbor embedding (t-SNE) [Van der Maaten/Hinton, 2008], or neighbor retrieval visualizer (NerV) [Venna et al., 2010] are used project data explicitly into two dimensions disregarding the subspaces of higher dimension than two and preserving only relevant neighborhoods in high-dimensional data. In the next step, the Delaunay graph [Delaunay, 1934] between the projected points is calculated, and each vertex between two projected points is weighted with the high-dimensional distance between the corresponding high-dimensional data points. Thereafter the shortest path between every pair of points is computed using the Dijkstra algorithm [Dijkstra, 1959]. The shortest paths are then used in the clustering process, which involves two choices depending on the structure type in the high-dimensional data [Thrun/Ultsch, 2020b]. This Boolean choice can be decided by looking at the topographic map of high-dimensional structures [Thrun/Ultsch, 2020a]. In a benchmarking of 34 comparable clustering methods, projection-based clustering was the only algorithm that always was able to find the high-dimensional distance or density-based structure of the dataset [Thrun/Ultsch, 2020b].

It should be noted that it is preferable to use a visualization for the Generalized U-Matrix like the topographic map plotTopographicMap of [Thrun et al., 2016] to evaluate the choice of the boolean parameter StructureType and the clustering, improve it or set the number of clusters appropriately. A comparison with 32 clustering algorithms showed that PBC is always able to find the correct cluster structure while the best of the 32 clustering algorithms varies depending on the dataset [Thrun/Ultsch, 2020].

The first systematic comparison to other DR clustering methods like Projection-Pursuit Methods ProjectionPursuitClustering, supspace clustering methods SubspaceClustering, and CA-based clustering methods can be found in [Thrun/Ultsch, 2020a]. For PCA-based clustering methods please see TandemClustering.

Value

List of

Cls	[1:n] numerical vector with n numbers defining the classification as the main output of the clustering algorithm. It has k unique numbers representing the arbitrary labels of the clustering. . Points which cannot be assigned to a cluster will be reported with 0.
Object	Object defined by clustering algorithm as the other output of this algorithm

Author(s)

Michael Thrun

References

[Bock, 1987] Bock, H.: On the interface between cluster analysis, principal component analysis, and multidimensional scaling, Multivariate statistical modeling and data analysis, (pp. 17-34), Springer, 1987.

[Thrun/Ultsch, 2017] Thrun, M. C., & Ultsch, A.: Projection based Clustering, Proc. International Federation of Classification Societies (IFCS), pp. 250-251, Tokai University, Japanese Classification Society (JCS), Tokyo, Japan August 7-10, 2017.

[Thrun/Ultsch, 2020a] Thrun, M. C., & Ultsch, A.: Using Projection based Clustering to Find Distance and Density based Clusters in High-Dimensional Data, Journal of Classification, in press, doi 10.1007/s00357-020-09373-2, 2020.

[Thrun et al., 2016] Thrun, M. C., Lerch, F., Loetsch, J., & Ultsch, A.: Visualization and 3D Printing of Multivariate Data of Biomarkers, in Skala, V. (Ed.), International Conference in Central Europe on Computer Graphics, Visualization and Computer Vision (WSCG), Vol. 24, pp. 7-16, Plzen, http://wscg.zcu.cz/wscg2016/short/A43-full.pdf, 2016.

[McInnes et al., 2018] McInnes, L., Healy, J., & Melville, J.: Umap: Uniform manifold approximation and projection for dimension reduction, arXiv preprint arXiv:1802.03426, 2018.

[Demartines/Herault, 1995] Demartines, P., & Herault, J.: CCA:" Curvilinear component analysis", Proc. 15 Colloque sur le traitement du signal et des images, Vol. 199, GRETSI, Groupe d Etudes du Traitement du Signal et des Images, France 18-21 September, 1995.

[Sammon, 1969] Sammon, J. W.: A nonlinear mapping for data structure analysis, IEEE Transactions on computers, Vol. 18(5), pp. 401-409. doi doi:10.1109/t-c.1969.222678, 1969.

[Thrun/Ultsch, 2020b] Thrun, M. C., & Ultsch, A.: Swarm Intelligence for Self-Organized Clustering, Journal of Artificial Intelligence, Vol. in press, pp. doi 10.1016/j.artint.2020.103237, 2020.

[Torgerson, 1952] Torgerson, W. S.: Multidimensional scaling: I. Theory and method, Psychometrika, Vol. 17(4), pp. 401-419. 1952.

[Venna et al., 2010] Venna, J., Peltonen, J., Nybo, K., Aidos, H., & Kaski, S.: Information retrieval perspective to nonlinear dimensionality reduction for data visualization, The Journal of Machine Learning Research, Vol. 11, pp. 451-490. 2010.

[Van der Maaten/Hinton, 2008] Van der Maaten, L., & Hinton, G.: Visualizing Data using t-SNE, Journal of Machine Learning Research, Vol. 9(11), pp. 2579-2605. 2008.

Examples

data('Hepta')
out=AutomaticProjectionBasedClustering(Hepta$Data,ClusterNo=7,PlotIt=FALSE)

FCPS documentation built on Oct. 19, 2023, 5:06 p.m.