Software Packages for Graphical Models / Bayesian Networks

Written by Kevin Murphy.
Last updated 31 October 2005.

Remarks

• A much more detailed comparison of some of these software packages isavailable from Appendix B of Bayesian AI, by Ann Nicholson and Kevin Korb.This appendix isavailablehere,and is based on the online comparison below.
• An online Frenchversion of this page is also available (not necessarily up-to-date).

What do the headers in the table mean?

• Src = source code included? (N=no) If so, what language?
• API = application program interface included?(N means the program cannot be integrated into your code, i.e., itmust be run as a standalone executable.)
• Exec = Executable runs on W = Windows (95/98/NT), U = Unix, M =Mac, or - = any machine with a compiler.
• Cts = are continuous (latent) nodes supported?G = (conditionally) Gaussians nodes supported analytically,Cs = continuous nodes supported by sampling,Cd = continuous nodes supported by discretization,Cx = continuous nodes supported by some unspecified method,D = only discrete nodes supported.
• GUI = Graphical User Interface included?
• Learns parameters?
• Learns structure? CI = means uses conditional independency tests
• Utility = utility and decision nodes (i.e., influence diagrams)supported?
• Free?0 = free (although possibly only for academic use).$ = commercial software (although most have free versionswhich are restricted invarious ways, e.g., the model size is limited, or models cannot besaved, or there is no API.)
• Undir?What kind of graphs are supported?U = only undirected graphs,D = only directed graphs,UD = both undirected and directed,CG = chain graphs (mixed directed/undirected).
• Inference = which inference algorithm is used?jtree = junction tree,varelim = variable (bucket) elimination,MH = Metropols Hastings,G = Gibbs sampling,IS = importance sampling,sampling = some other Monte Carlo method,polytree = Pearl's algorithm restricted to a graph with no cycles, none = no inference supported (hence the program is only designed forstructure learning from completely observed data)
• Comments.If in "quotes", I am quoting the authors at their request.

Name

Authors

Src

API

Exec

Cts

GUI

Params

Struct

Utility

Free

Undir

Inference

Comments

AgenaRisk

Agena

N

Y

W,U

Cx

Y

Y

N

N

$

D

JTree

Simulation by Dynamic discretisation

Analytica

Lumina

N

Y

W,M

G

Y

N

N

Y

$

D

sampling

spread sheet compatible

Banjo

Hartemink

Java

Y

W,U,M

Cd

N

N

Y

N

0

D

none

structure learning ofstatic or dynamic networks of discrete variables

Bassist

U. Helsinki

C++

Y

U

G

N

Y

N

N

0

D

MH

Generates C++ for MCMC.

Bayda

U. Helsinki

Java

Y

WUM

G

Y

Y

N

N

0

D

?

Bayesian Naive Bayes classifier.

BayesBuilder

Nijman (U. Nijmegen)

N

N

W

D

Y

N

N

N

0

D

?

-

BayesiaLab

Bayesia Ltd

N

N

-

Cd

Y

Y

Y

N

$

CG

jtree,G

Structural learning, adaptivequestionnaires, dynamic models

Bayesware Discoverer

Bayesware

N

N

WUM

Cd

Y

Y

Y

N

$

D

?

Uses bound and collapse for learning with missing data.

B-course

U. Helsinki

N

N

WUM

Cd

Y

Y

Y

N

0

D

?

Runs on their server: view results using a web browser.

Belief net power constructor

Cheng (U.Alberta)

N

W

W

D

Y

Y

CI

N

0

D

?

-

BNT

Murphy (U.C.Berkeley)

Matlab/C

Y

WUM

G

N

Y

Y

Y

0

D,U

Many

Also handles dynamic models, like HMMs and Kalman filters.

BNJ

Hsu (Kansas)

Java

-

-

D

Y

N

Y

N

0

D

jtree, IS

-

BucketElim

Rish (U.C.Irvine)

C++

Y

WU

D

N

N

N

N

0

D

Varelim

-

BUGS

MRC/Imperial College

N

N

WU

Cs

W

Y

N

N

0

D

Gibbs

-

Business Navigator 5

Data Digest Corp

N

N

W

Cd

Y

Y

Y

N

$

D

Jtree

-

CABeN

Cousins et al. (Wash. U.)

C

Y

WU

D

N

N

N

N

0

D

5 Sampling methods

-

Causal discoverer

Vanderbilt

N

N

W

-

-

N

Y

N

0

D

-

structure learning only

CoCo+Xlisp

Badsberg (U. Aalborg)

C/lisp

Y

U

D

Y

Y

CI

N

0

U

Jtree

Designed for contingency tables.

CIspace

Poole et al. (UBC)

Java

N

WU

D

Y

N

N

N

0

D

Varelim

-

DBNbox

Roberts et al

Matlab

-

-

Y

N

Y

N

N

Y

D

Various

DBNs

Deal

Bottcher et al

R

-

-

G

Y

Y

Y

N

0

D

None

Structure learning.

DeriveIt

DeriveIt LLC

N

-

-

?

?

Y

Y

?

$

D

Jtree

Exploits local structure in CPDs.

Ergo

Noetic systems

N

Y

W,M

D

Y

N

N

N

$

D

jtree

-

GDAGsim

Wilkinson (U. Newcastle)

C

Y

WUM

G

N

N

N

N

0

D

Exact

Bayesian analysis of large linear Gaussian directed models.

Genie

U. Pittsburgh

N

WU

WU

D

W

N

N

Y

0

D

Jtree

-

GMRFsim

Rue (U. Trondheim)

C

Y

WUM

G

N

N

N

N

0

U

MCMC

Bayesian analysis of large linear Gaussian undirected models.

GMTk

Bilmes (UW), Zweig (IBM)

N

Y

U

D

N

Y

Y

N

0

D

Jtree

Designed for speech recognition.

gR

Lauritzen et al.

R

-

-

-

-

-

-

-

0

-

-

Currently vaporware

Grappa

Green (Bristol)

R

-

-

D

N

N)

N

N

0

D

Jtree

-

Hugin Expert

Hugin

N

Y

W

G

W

Y

CI

Y

$

CG

Jtree

-

Hydra

Warnes (U.Wash.)

Java

-

-

Cs

Y

Y

N

N

0

U,D

MCMC

-

Ideal

Rockwell

Lisp

Y

WUM

D

Y

N

N

Y

0

D

Jtree

GUI requires Allegro Lisp.

Java Bayes

Cozman (CMU)

Java

Y

WUM

D

Y

N

N

Y

0

D

Varelim, jtree

-

KBaseAI

Codeas

N

Y

W,U

D

N

N

N

N

$

D

varelim

client/server architecture, multiple users, accesscontrol, query language

LibB

Friedman (Hebrew U)

N

Y

W

D

N

Y

Y

N

0

D

none

Structure learning

MIM

HyperGraph Software

N

N

W

G

Y

Y

Y

N

$

CG

Jtree

Up to 52 variables.

MSBNx

Microsoft

N

Y

W

D

W

N

N

Y

0

D

Jtree

-

Netica

Norsys

N

WUM

W

G

W

Y

N

Y

$

D

jtree

-

OptimalReinsertion

Moore, Wong (CMU)

N

N

W,U

D

N

Y

Y

N

0

D

none

structure learning

PMT

Pavlovic (BU)

Matlab/C

-

-

D

N

Y

N

N

0

D

special purpose

-

PNL

Eruhimov (Intel)

C++

-

-

D

N

Y

Y

N

0

U,D

Jtree

A C++ version of BNT; will be released 12/03.

Pulcinella

IRIDIA

Lisp

Y

WUM

D

Y

N

N

N

0

D

?

Uses valuation systems for non-probabilistic calculi.

RISO

Dodier (U.Colorado)

Java

Y

WUM

G

Y

N

N

N

0

D

Polytree

Distributed implementation.

Sam Iam

Darwiche (UCLA)

N

N ?

WU ? (Java executable)

G ?

Y

Y

N ?

Y

0

D

Recursive conditioning

Also does sensitivity Analysis

Tetrad

CMU

N

N

WU

G

N

Y

CI

N

0

U,D

None

-

UnBBayes

?

Java

-

-

D

Y

N

Y

N

0

D

jtree

K2 for struct learning

Vibes

Winn & Bishop (U. Cambridge)

Java

Y

WU

Cx

Y

Y

N

N

0

D

Variational

Not yet available.

Web Weaver

Xiang (U.Regina)

Java

Y

WUM

D

Y

N

N

Y

0

D

?

-

WinMine

Microsoft

N

N

W

Cx

Y

Y

Y

N

0

U,D

None

Learns BN or dependency net structure.

XBAIES 2.0

Cowell (City U.)

N

N

W

G

Y

Y

N

Y

0

CG

Jtree

-

Other sites related to (software for) graphical models

• Google's list of Bayes net software.
• Comparison of decision analysis software packages byH錵an L. Younes. The emphasis is on real-time decision making.
• Older belief net programs (c 1996), a list created (but no longer maintained)by Russ Almond.
• Repository of Bayes nets.