ICL (Integrated Complete-data Likelihood) is a model selection criterion
that penalizes uncertain cluster assignments more heavily than BIC.
This makes it more conservative for selecting the number of clusters K.
Usage
compute_ICL_mpca(loglik, K, r, M, N_obs, resp)
Arguments
- loglik
The total log-likelihood from the MPCA model.
- K
Number of clusters.
- r
Principal component dimension.
- M
Observed dimension (channels).
- N_obs
Number of observations (typically sum of all T_i across subjects).
- resp
An (N x K) matrix of responsibilities (posterior probabilities).
Value
A numeric ICL value.
Details
ICL is computed as:
$$ICL = BIC + 2 \times Entropy$$
where Entropy = -sum(resp * log(resp)).