integrate_similarity_matrices {Spectrum}

integrate_similarity_matrices: integrate similarity matrices using a tensor product graph linear combination and diffusion technique

Description

Given a list of similarity matrices this function will integrate them running the Shu algorithm, also can reduce noise if the input is a list consisting of a single matrix.

Usage

integrate_similarity_matrices(kernellist, KNNs_p = 10,
  diffusion_iters = 4, method = "TPG")

Arguments

kernellist

A list of similarity matrices: those to be integrated

KNNs_p

Numerical value: number of nearest neighbours for KNN graph (default=10, suggested=10-20)

diffusion_iters

Numerical value: number of iterations for graph diffusion (default=4, suggested=2-6)

method

Character: either TPG (see reference below) or mean (default=TPG)

Value

An integrated similarity matrix

References

Shu, Le, and Longin Jan Latecki. "Integration of single-view graphs with diffusion of tensor product graphs for multi-view spectral clustering." Asian Conference on Machine Learning. 2016.

Examples

i_test <- integrate_similarity_matrices(misslfilled,method='mean')