BIC.threshold: BIC threshold for final parameter estimation

In LinearDetect: Change Point Detection in High-Dimensional Linear Regression Models

Description

BIC threshold for final parameter estimation

Usage

BIC.threshold(
  method,
  beta.final,
  k,
  m.hat,
  brk,
  data_y,
  data_x = NULL,
  b_n = 2,
  nlam = 20
)

Arguments

method	method name for the model: Constant: Mean-shift Model; MvLR: Multivariate Linear Regression; MLR: Multiple Linear Regression
beta.final	a combined matrix of estimated parameter coefficient matrices for all stationary segementations
k	dimensions of parameter coefficient matrices
m.hat	number of estimated change points
brk	vector of estimated change points
data_y	input data matrix (response), with each column representing the time series component
data_x	input data matrix (predictor), with each column representing the time series component
b_n	the block size
nlam	number of hyperparameters for grid search

Value

lambda.val.best, the tuning parameter lambda selected by BIC.

LinearDetect documentation built on March 22, 2021, 9:06 a.m.