bic: Bayesian Information Criterion (BIC) value.

In TensorComplete: Tensor Noise Reduction and Completion Methods

Bayesian Information Criterion (BIC) value.

Description

Compute Bayesian Information Criterion (BIC) given a parameter tensor, an observed tensor, the dimension, and the rank based on cumulative logistic model. This BIC function is designed for selecting rank in the fit_ordinal function.

Usage

bic(ttnsr,theta,omega,d,r)

Arguments

ttnsr	An observed tensor.
theta	A continuous-valued tensor (latent parameters).
omega	The cut-off points.
d	Dimension of the tensor.
r	Rank of the tensor.

Value

BIC value at given inputs based on cumulative logistic model.

TensorComplete documentation built on April 14, 2023, 9:10 a.m.