MMHC (bnlearn)

Module name	bnlearn_mmhc
Package	bnlearn
Version	4.8.3
Language	R
Docs	here
Paper	Tsamardinos et al.[1]
Graph type	DAG
MCMC	No
Edge constraints	Yes
Data type	C, D, M
Data missingness
Intervention type
Docker	bpimages/bnlearn:4.8.3

Max-Min Hill-Climbing

Max-min hill-climbing (MMHC) is a hybrid method which first estimates the skeleton of a DAG using an algorithm called Max-Min Parents and Children and then performs a greedy hill-climbing search to orient the edges with respect to a Bayesian score. It is a popular approach used as standard benchmark and also well suited for high- dimensional domains.

Some fields described

edgeConstraints Name of the JSON file containing background knowledge

Example JSON

[
  {
    "id": "mmhc-bde-mi",
    "alpha": [
      0.01,
      0.05,
      0.1
    ],
    "test": "mi",
    "score": "bde",
    "iss": 0.1,
    "issmu": 1,
    "issw": null,
    "l": 5,
    "k": 1,
    "prior": "uniform",
    "beta": 1,
    "timeout": null,
    "edgeConstraints": "edgeConstraints.json"
  },
  {
    "id": "mmhc-bge-zf",
    "alpha": [
      0.001,
      0.01,
      0.05,
      0.1
    ],
    "score": "bge",
    "test": "zf",
    "iss": 1,
    "issmu": 1,
    "issw": null,
    "l": 5,
    "k": 1,
    "prior": "uniform",
    "beta": null,
    "timeout": null,
    "edgeConstraints": "edgeConstraints.json"
  }
]