hybrid: Hybrid structure learning algorithms

In vspinu/bnlearn: Bayesian network structure learning, parameter learning and inference

Description

Learn the structure of a Bayesian network with the Max-Min Hill Climbing (MMHC) and the more general 2-phase Restricted Maximization (RSMAX2) hybrid algorithms.

Usage

rsmax2(x, whitelist = NULL, blacklist = NULL, restrict, maximize = "hc",
  test = NULL, score = NULL, alpha = 0.05, B = NULL, ...,
  maximize.args = list(), optimized = TRUE, strict = FALSE, debug = FALSE)
mmhc(x, whitelist = NULL, blacklist = NULL, test = NULL, score = NULL,
  alpha = 0.05, B = NULL, ..., restart = 0, perturb = 1, max.iter = Inf,
  optimized = TRUE, strict = FALSE, debug = FALSE)

Arguments

x	a data frame containing the variables in the model.
whitelist	a data frame with two columns (optionally labeled "from" and "to"), containing a set of arcs to be included in the graph.
blacklist	a data frame with two columns (optionally labeled "from" and "to"), containing a set of arcs not to be included in the graph.
restrict	a character string, the constraint-based algorithm to be used in the “restrict” phase. Possible values are gs, iamb, fast.iamb, inter.iamb and mmpc. See bnlearn-package and the documentation of each algorithm for details.
maximize	a character string, the score-based algorithm to be used in the “maximize” phase. Possible values are hc and tabu. See bnlearn-package for details.
test	a character string, the label of the conditional independence test to be used by the constraint-based algorithm. If none is specified, the default test statistic is the mutual information for categorical variables, the Jonckheere-Terpstra test for ordered factors and the linear correlation for continuous variables. See bnlearn-package for details.
score	a character string, the label of the network score to be used in the score-based algorithm. If none is specified, the default score is the Bayesian Information Criterion for both discrete and continuous data sets. See bnlearn-package for details.
alpha	a numeric value, the target nominal type I error rate of the conditional independence test.
B	a positive integer, the number of permutations considered for each permutation test. It will be ignored with a warning if the conditional independence test specified by the test argument is not a permutation test.
...	additional tuning parameters for the network score used by the score-based algorithm. See score for details.
maximize.args	a list of arguments to be passed to the score-based algorithm specified by maximize, such as restart for hill-climbing or tabu for tabu search.
restart	an integer, the number of random restarts for the score-based algorithm.
perturb	an integer, the number of attempts to randomly insert/remove/reverse an arc on every random restart.
max.iter	an integer, the maximum number of iterations for the score-based algorithm.
debug	a boolean value. If TRUE a lot of debugging output is printed; otherwise the function is completely silent.
optimized	a boolean value. See bnlearn-package for details.
strict	a boolean value. If TRUE conflicting results in the learning process generate an error; otherwise they result in a warning.

Value

An object of class bn. See bn-class for details.

Note

mmhc is simply rshc with restrict set to mmpc and maximize set to hc.

Author(s)

Marco Scutari

References

Tsamardinos I, Brown LE, Aliferis CF (2006). "The Max-Min Hill-Climbing Bayesian Network Structure Learning Algorithm". Machine Learning, 65(1), 31-78.

See Also

local discovery algorithms, score-based algorithms, constraint-based algorithms.

vspinu/bnlearn documentation built on May 3, 2019, 7:08 p.m.