v2

latestOpenAPI 3.0.1Apache 2.02026-07-2623201.3 MB

Assets / Clustering / Correlation-Based

Spectral Clustering

Compute a partition of a universe of assets into clusters using a spectral clustering method.

References

Brownlees, C., Guomundsson, G. S., & Lugosi, G. (2020). Community Detection in Partial Correlation Network Models. Journal of Business & Economic Statistics, 40(1), 216–226
Mihai Cucuringu, Peter Davies, Aldo Glielmo, Hemant Tyagi, SPONGE: A generalized eigenproblem for clustering signed networks, Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics, PMLR 89:1088-1098, 2019

post/assets/clustering/spectral/correlation-based

Request body

assetsinteger required

The number of assets

clusteringMethod'blockbuster' | 'sponge' | 'spongeSymmetric'

The hierarchical clustering method to use

clustersinteger

The number of clusters to group the assets into; if not provided, it is computed using a proprietary procedure

clusteringSimulationsinteger

The number of clustering simulations to perform when the number of clusters is computed using a proprietary variation of Horn's parallel analysis method

positiveDegreeMatrixRegularizationParameternumber

The regularization parameter applicable to the positive degree matrix, in case the clustering method is either SPONGE or symmetric SPONGE

negativeDegreeMatrixRegularizationParameternumber

The regularization parameter applicable to the positive degree matrix, in case the clustering method is either SPONGE or symmetric SPONGE

v2

Spectral Clustering

Request body

Response