EM_RES: EM algorithm for mixture of RES distributions defined by g,...

In Mufabo/ICASSP20.T6.R: ICASSP20.T6.R

Description

EM algorithm for mixture of RES distributions defined by g, psi

Usage

EM_RES(data, ll, g, psi, limit = 1e-06, em_max_iter = 200,
  reg_value = 1e-06, test_args = NULL)

Arguments

data	Matrix[N, r] data without labels
ll	scalar, number of clusters
g	function(t), gauss
reg_value

Value

list

1. mu_hat matrix[r, ll] Estimate of cluster centers

2. S_hat array[r, r, ll] Estimate of cluster scatter matrices

3. t matrix[N, ll] Squared Mahalanobis distances of each point to each cluster

4. R matrix[N, ll] Estimate of the posterior probabilites per cluster.

References

1. F. K. Teklehaymanot, M. Muma, and A. M. Zoubir, "Bayesian Cluster Enumeration Criterion for Unsupervised Learning", IEEE Trans. Signal Process. (accepted), [Online-Edition: https://arxiv.org/abs/1710.07954v2], 2018.

2. F. K. Teklehaymanot, M. Muma, and A. M. Zoubir, "Novel Bayesian Cluster Enumeration Criterion for Cluster Analysis With Finite Sample Penalty Term", in Proc. 43rd IEEE Int. conf. on Acoustics, Speech and Signal Process. (ICASSP), pp. 4274-4278, 2018, [Online-Edition: https://www.researchgate.net/publication/322918028]

Mufabo/ICASSP20.T6.R documentation built on May 30, 2021, 11:20 a.m.