Truncating Regular Vine Copula Based on Mutual Information: An Efficient Parsimonious Model for High-Dimensional Data

Table 3

Selection criteria and log-likelihood values of the proposed truncation method and the sequential truncation method of [11] (no pruning pairs).

Name	Level	No. of parameters	Log-likelihood	AIC
Full specification model	14	105	5043.277	−9834.554
MI method (new)	3	39	4780.808	−9425.617
Sequential [11]	6	69	4924.329	−9652.657