SCCI (version 1.2)

regret: Multinomial Regret Term

Description

Calculates the multinomial regret term for for a discrete random variable with domain size \(k\) and sample size \(n\) (see Silander et al. 2018). Note that we use the logarithm to basis 2 to calculate the result. To compare the results to Silander et al. (2018), we need to multiply the result with \(log(2)\) to compare the results.

Usage

regret(n,k)

Arguments

n

Integer (sample size)

k

Integer (domain size)

References

Tomi Silander, Janne Lepp<U+00E4>-aho, Elias J<U+00E4><U+00E4>saari, Teemu Roos; Quotient normalized maximum likelihood criterion for learning bayesian network structures, Proceedings of the 21nd International Conference on Artificial Intelligence and Statistics (AISTATS), PMLR, 2018

Examples

# NOT RUN {
regret(50,10)           ## 19.1
regret(50,10) * log(2)  ## 13.24 (see Silander et al. 2018)
# }