CPD_DLM: Setting up the CPD_DLM model

In SKFCPD: Fast Online Changepoint Detection for Temporally Correlated Data

Setting up the CPD_DLM model

Description

Implementing the robust GaSP model for estimating the changepoint locations. The range parameter and noise-to-signal ratio are estimated from the training samples by a Gaussian process model.

Usage

  CPD_DLM(design, response, gamma,model_type, mu, sigma_2, eta,
         kernel_type, stop_at_first_cp, hazard_vec,
         truncate_at_prev_cp)

Arguments

design	A matrix with dimension n x p. The design of the experiment.
response	A matrix with dimension n x q. The observations.
gamma	A numeric variable of the range parameter for the covariance matrix. The default value of gamma is 1.
model_type	A numeric variable that can take values of 0, 1 and 2. Model_type=0 stands for a GP model with unknown mean and known variance. Model_type=1 stands for a GP model with known mean and unknown variance. Model_type=2 stands for a GP model with unknown mean and unknown variance. The default value of model_type is 2.
mu	A vector of the mean parameter at each coordinate. Ignored when model_type = 0 or 2.
sigma_2	A vector of the variance parameter at each coordinate.
eta	A vector of the noise-to-signal ratio at each coordinate
kernel_type	A character specifying the type of kernels of the input. matern_5_2 are Matern correlation with roughness parameter 5/2. exp is power exponential correlation with roughness parameter alpha=2. The default choice is matern_5_2.
stop_at_first_cp	A numeric variable that decides if the SKFCPD method stops when it detects the first changepoint. The default value of stop_at_first_cp is FALSE.
hazard_vec	The hazard vector in the SKFCPD method. 1/vector is the prior probability that a changepoint occur at a vector of time points.
truncate_at_prev_cp	If TRUE, truncate the run length at the most recently detected changepoint. The default value of truncate_at_prev_cp is FALSE.

Value

SKFCPD returns a S4 object of class SKFCPD (see SKFCPD-class).

Author(s)

Hanmo Li [aut, cre], Yuedong Wang [aut], Mengyang Gu [aut]

Maintainer: Hanmo Li <[e-mail]>

References

Li, Hanmo, Yuedong Wang, and Mengyang Gu. Sequential Kalman filter for fast online changepoint detection in longitudinal health records. arXiv preprint arXiv:2310.18611 (2023).

Fearnhead, Paul, and Zhen Liu. On-line inference for multiple changepoint problems. Journal of the Royal Statistical Society Series B: Statistical Methodology 69, no. 4 (2007): 589-605.

Adams, Ryan Prescott, and David JC MacKay. Bayesian online changepoint detection. arXiv preprint arXiv:0710.3742 (2007).

Hartikainen, Jouni, and Simo Sarkka. Kalman filtering and smoothing solutions to temporal Gaussian process regression models. In 2010 IEEE international workshop on machine learning for signal processing, pp. 379-384. IEEE, 2010.

SKFCPD documentation built on June 22, 2024, 11:06 a.m.