R-LAIR: Riverside Lab for Artificial Intelligence Research
Continuous Time Bayesian Network Reasoning and Learning Engine (CTBN-RLE)
v1.0.2: CTBN papers
The following are the academic papers we know about that pertain to
continuous time Bayesian networks. The code assumes knowledge of at least
the first paper which is included in this directory.
BASIC FORMULATION:
Uri Nodelman, Christian R. Shelton, and Daphne Koller
(2002). "Continuous Time Bayesian Networks." Proceedings
of the Eighteenth International Conference on
Uncertainty in Artificial Intelligence (pp. 378-387).
http://www.cs.ucr.edu/~cshelton/papers/index.cgi?NodSheKol02
[First paper formulating CTBNs. It defines a CTBN's syntax and semantics
and briefly discusses an approximate inference method.]
Tal El-Hay, Nir Friedman, Daphne Koller and Raz Kupferman (2006).
"Continuous Time Markov Networks", Proceedings of the Twenty-Second
Conference on Uncertainty in Artificial Intelligence (pp. 155-164).
http://ai.stanford.edu/~koller/Papers/El-Hay+al:UAI06.pdf
[An undirected version of CTBNs that represents the subset of reversable
Markov processes.]
LEARNING:
Uri Nodelman, Christian R. Shelton, and Daphne Koller (2003). "Learning
Continuous Time Bayesian Networks." Proceedings of the Nineteenth
International Conference on Uncertainty in Artificial Intelligence
(pp. 451-458).
http://www.cs.ucr.edu/~cshelton/papers/index.cgi?NodSheKol03
[Introduces the sufficient statistics for a CTBN and formulates the
maximum likelihood and maximum a posteriori parameter learning formulas
for the complete data case. It also formulates a structure learning
method based on a Bayesian scoring function.]
Uri Nodelman, Christian R. Shelton, and Daphne Koller (2005). "Expectation
Maximization and Complex Duration Distributions for Continuous Time
Bayesian Networks." Proceedings of the Twenty-First International
Conference on Uncertainty in Artificial Intelligence (pp. 421-430).
http://www.cs.ucr.edu/~cshelton/papers/index.cgi?NodSheKol05
[Introduces expectation maximization for CTBNs to extend the parameter
and structure learning methods of the previous paper to the incomplete
data case.]
INFERENCE:
(The first paper from 2002 also discusses this to some degree)
Uri Nodelman, Daphne Koller, and Christian R. Shelton (2005). "Expectation
Propagation for Continuous Time Bayesian Networks." Proceedings of the
Twenty-First International Conference on Uncertainty in Artificial
Intelligence (pp. 431-440).
http://www.cs.ucr.edu/~cshelton/papers/index.cgi?NodKolShe05
[An approximate inference method based on expectation propagation.]
Brenda Ng, Avi Pfeffer, and Richard Dearden (2005). "Continuous Time
Particle Filtering." Proceedings of the Nineteenth International
Joint Conference on Artificial Intelligence (pp. 1360-1365).
[A particle filter for the case of an unobserved CTBN that is driving
stchastic differential equations whose values are observed at
discrete time points.]
Suchi Saria, Uri Nodelman, and Daphne Koller (2007). "Reasoning at the
Right Time Granularity." Proceedings of the Twenty-third Conference
on Uncertainty in Artificial Intelligence" (pp. 421-430).
http://ai.stanford.edu/~nodelman/papers/dynamic-EP.pdf
[An improved approximate inference method based on expectation
propagation.]
Yu Fan and Christian R. Shelton (2008). "Sampling for Approximate
Inference in Continuous Time Bayesian Networks." Proceedings of
the Tenth International Symposium on Artificial Intelligence
and Mathematics.
http://www.cs.ucr.edu/~cshelton/papers/index.cgi?FanShe08
[An importance sampling method and its extension to particle filtering
and smoothing.]
Tal El-Hay, Nir Friedman, and Raz Kupferman (2008). "Gibbs Sampling in
Factorized Continuous-Time Markov Processes." Proceedings of the
Twenty-Fourth Conference on Uncertainty in Artificial Intelligence
(pp. 169-178).
http://uai2008.cs.helsinki.fi/UAI_camera_ready/el-hay.pdf
[A method for Gibbs sampling for a CTBN.]
Ido Cohn, Tal El-Hay, Raz Kupferman, and Nir Friedman (2009). "Mean
Field Variational Approximation for Continuous-Time Bayesian
Networks." Proceedings of the Twenty-Fifth International
Conference on Uncertainty in Artificial Intelligence.
http://www.cs.mcgill.ca/~uai2009/papers/
UAI2009_0134_c9c16e478d8e82ecc3848c5dc76b4925.pdf
[A mean field variational method for a CTBN.]
OTHER EXTENSIONS:
Karthik Gopalratnam, Henry Kautz, and Daniel S. Weld (2005). "Extending
Continuous Time Bayesian Networks." Proceedings of 20th National
Conference on Artificial Intelligence-AAAI 2005 (pp. 981-986).
http://www.cs.washington.edu/homes/weld/papers/aaai05_CTBN.pdf
[This presents an alternative to exponential duration distributions.]
Kin Fai Kan and Christian R. Shelton (2008). "Solving Structured
Continuous-Time Markov Decision Processes." Tenth International
Symposium on Artificial Intelligence and Mathematics.
http://www.cs.ucr.edu/~cshelton/papers/index.cgi?KanShe08
[This presents an approximate continuous Markov decision process planning
solution using CTBN.]
Luigi Portinale and Daniele Codetta-Raiteri (2009). "Generalizing
Continuous Time Bayesian Networks with Immediate Nodes."
Proceedings on the Workshop on Graph Structured for Knowledge
Representation and Reasoning.
http://web.unipmn.it/%7Eraiteri/papers/gkr.pdf
[Presents an extension for describing simultaneous transitions in a
CTBN and links it to generalized stochastic Petri nets.]
APPLICATIONS:
(most of the papers above have at least one application of a CTBN. These
papers' main contributions are in applications, rather than theory.)
Uri Nodelman and Eric Horvitz (2003). "Continuous Time Bayesian Networks
for Inferring Users' Presence and Activities with Extensions for
Modeling and Evaluation." MSR-TR-2003-97, Microsoft Research.
http://research.microsoft.com/en-us/um/people/horvitz/uri_eh.pdf
[Applies CTBNs to modeling user's use of desktop applications.]
Ralf Herbrich, Thore Graepel, and Brendan Murphy (2007). "Structure from
Failure." SYSML '07: Proceedings of the 2nd USENIX workshop on tackling
computer system problems with machine learning. pp. 1-6.
http://www.usenix.org/event/sysml07/tech/full_papers/herbrich/herbrich.pdf
[Applies CTBNs to modeling failures in server farms.]
Jing Xu and Christian R. Shelton (2008). "Continuous Time Bayesian Networks
for Host Level Network Intrusion Detection." Machine Learning
and Knowledge Discovery in Databases (ECML/PKDD) (LNAI, vol 5212)
(pp. 613-627).
http://www.cs.ucr.edu/~cshelton/papers/index.cgi?XuShe08
[Applies CTBNs to host-level computer network intrusion detection.
Introduces "toggle" variables.]
Yu Fan and Christian R. Shelton (2009). "Learning Continuous-Time Social
Network Dynamics." Proceedings of the Twenty-Fifth International Conference
on Uncertainty in Artificial Intelligence.
http://www.cs.ucr.edu/~cshelton/papers/index.cgi?YuShe09
[Applies CTBNs to social network dynamics. Introduces MCMC inference
method.]
Continuous Time Bayesian Network Reasoning and Learning Engine (CTBN-RLE)
v1.0.2: CTBN papers
The following are the academic papers we know about that pertain to continuous time Bayesian networks. The code assumes knowledge of at least the first paper which is included in this directory. BASIC FORMULATION: Uri Nodelman, Christian R. Shelton, and Daphne Koller (2002). "Continuous Time Bayesian Networks." Proceedings of the Eighteenth International Conference on Uncertainty in Artificial Intelligence (pp. 378-387). http://www.cs.ucr.edu/~cshelton/papers/index.cgi?NodSheKol02 [First paper formulating CTBNs. It defines a CTBN's syntax and semantics and briefly discusses an approximate inference method.] Tal El-Hay, Nir Friedman, Daphne Koller and Raz Kupferman (2006). "Continuous Time Markov Networks", Proceedings of the Twenty-Second Conference on Uncertainty in Artificial Intelligence (pp. 155-164). http://ai.stanford.edu/~koller/Papers/El-Hay+al:UAI06.pdf [An undirected version of CTBNs that represents the subset of reversable Markov processes.] LEARNING: Uri Nodelman, Christian R. Shelton, and Daphne Koller (2003). "Learning Continuous Time Bayesian Networks." Proceedings of the Nineteenth International Conference on Uncertainty in Artificial Intelligence (pp. 451-458). http://www.cs.ucr.edu/~cshelton/papers/index.cgi?NodSheKol03 [Introduces the sufficient statistics for a CTBN and formulates the maximum likelihood and maximum a posteriori parameter learning formulas for the complete data case. It also formulates a structure learning method based on a Bayesian scoring function.] Uri Nodelman, Christian R. Shelton, and Daphne Koller (2005). "Expectation Maximization and Complex Duration Distributions for Continuous Time Bayesian Networks." Proceedings of the Twenty-First International Conference on Uncertainty in Artificial Intelligence (pp. 421-430). http://www.cs.ucr.edu/~cshelton/papers/index.cgi?NodSheKol05 [Introduces expectation maximization for CTBNs to extend the parameter and structure learning methods of the previous paper to the incomplete data case.] INFERENCE: (The first paper from 2002 also discusses this to some degree) Uri Nodelman, Daphne Koller, and Christian R. Shelton (2005). "Expectation Propagation for Continuous Time Bayesian Networks." Proceedings of the Twenty-First International Conference on Uncertainty in Artificial Intelligence (pp. 431-440). http://www.cs.ucr.edu/~cshelton/papers/index.cgi?NodKolShe05 [An approximate inference method based on expectation propagation.] Brenda Ng, Avi Pfeffer, and Richard Dearden (2005). "Continuous Time Particle Filtering." Proceedings of the Nineteenth International Joint Conference on Artificial Intelligence (pp. 1360-1365). [A particle filter for the case of an unobserved CTBN that is driving stchastic differential equations whose values are observed at discrete time points.] Suchi Saria, Uri Nodelman, and Daphne Koller (2007). "Reasoning at the Right Time Granularity." Proceedings of the Twenty-third Conference on Uncertainty in Artificial Intelligence" (pp. 421-430). http://ai.stanford.edu/~nodelman/papers/dynamic-EP.pdf [An improved approximate inference method based on expectation propagation.] Yu Fan and Christian R. Shelton (2008). "Sampling for Approximate Inference in Continuous Time Bayesian Networks." Proceedings of the Tenth International Symposium on Artificial Intelligence and Mathematics. http://www.cs.ucr.edu/~cshelton/papers/index.cgi?FanShe08 [An importance sampling method and its extension to particle filtering and smoothing.] Tal El-Hay, Nir Friedman, and Raz Kupferman (2008). "Gibbs Sampling in Factorized Continuous-Time Markov Processes." Proceedings of the Twenty-Fourth Conference on Uncertainty in Artificial Intelligence (pp. 169-178). http://uai2008.cs.helsinki.fi/UAI_camera_ready/el-hay.pdf [A method for Gibbs sampling for a CTBN.] Ido Cohn, Tal El-Hay, Raz Kupferman, and Nir Friedman (2009). "Mean Field Variational Approximation for Continuous-Time Bayesian Networks." Proceedings of the Twenty-Fifth International Conference on Uncertainty in Artificial Intelligence. http://www.cs.mcgill.ca/~uai2009/papers/ UAI2009_0134_c9c16e478d8e82ecc3848c5dc76b4925.pdf [A mean field variational method for a CTBN.] OTHER EXTENSIONS: Karthik Gopalratnam, Henry Kautz, and Daniel S. Weld (2005). "Extending Continuous Time Bayesian Networks." Proceedings of 20th National Conference on Artificial Intelligence-AAAI 2005 (pp. 981-986). http://www.cs.washington.edu/homes/weld/papers/aaai05_CTBN.pdf [This presents an alternative to exponential duration distributions.] Kin Fai Kan and Christian R. Shelton (2008). "Solving Structured Continuous-Time Markov Decision Processes." Tenth International Symposium on Artificial Intelligence and Mathematics. http://www.cs.ucr.edu/~cshelton/papers/index.cgi?KanShe08 [This presents an approximate continuous Markov decision process planning solution using CTBN.] Luigi Portinale and Daniele Codetta-Raiteri (2009). "Generalizing Continuous Time Bayesian Networks with Immediate Nodes." Proceedings on the Workshop on Graph Structured for Knowledge Representation and Reasoning. http://web.unipmn.it/%7Eraiteri/papers/gkr.pdf [Presents an extension for describing simultaneous transitions in a CTBN and links it to generalized stochastic Petri nets.] APPLICATIONS: (most of the papers above have at least one application of a CTBN. These papers' main contributions are in applications, rather than theory.) Uri Nodelman and Eric Horvitz (2003). "Continuous Time Bayesian Networks for Inferring Users' Presence and Activities with Extensions for Modeling and Evaluation." MSR-TR-2003-97, Microsoft Research. http://research.microsoft.com/en-us/um/people/horvitz/uri_eh.pdf [Applies CTBNs to modeling user's use of desktop applications.] Ralf Herbrich, Thore Graepel, and Brendan Murphy (2007). "Structure from Failure." SYSML '07: Proceedings of the 2nd USENIX workshop on tackling computer system problems with machine learning. pp. 1-6. http://www.usenix.org/event/sysml07/tech/full_papers/herbrich/herbrich.pdf [Applies CTBNs to modeling failures in server farms.] Jing Xu and Christian R. Shelton (2008). "Continuous Time Bayesian Networks for Host Level Network Intrusion Detection." Machine Learning and Knowledge Discovery in Databases (ECML/PKDD) (LNAI, vol 5212) (pp. 613-627). http://www.cs.ucr.edu/~cshelton/papers/index.cgi?XuShe08 [Applies CTBNs to host-level computer network intrusion detection. Introduces "toggle" variables.] Yu Fan and Christian R. Shelton (2009). "Learning Continuous-Time Social Network Dynamics." Proceedings of the Twenty-Fifth International Conference on Uncertainty in Artificial Intelligence. http://www.cs.ucr.edu/~cshelton/papers/index.cgi?YuShe09 [Applies CTBNs to social network dynamics. Introduces MCMC inference method.]
