Probabilistic network models

Blau et al. [22] have applied probabilistic network models to model resource needs and success probabilities in pharmaceutical and agrochemical development, through Monte Carlo analysis. This requires solving the problem of scheduling a portfolio of projects under uncertainty about progression. This approach is tractable for drug development. However, the inherent complex-... 

A. Frank and P. Pevzner. PepNovo De novo Peptide Sequencing via Probabilistic Network Modeling. Anal. Chem., 77, no. 4 (2005) 964-973. 

In the near future, more sophisticated models can be built using probabilistic networks. A probabilistic network is a factorization of the joint probabiHty function over all the considered variables (markers, interventions, and outcomes) based on knowledge about the dependencies and independencies between the variables. Such knowledge is naturally provided by the hits coming out of the association screen, where each association can be interpreted as a dependency, and the absence of an association as an independency between variables. The model can then be parameterized by fitting to the data, similarly to the linear and logistic regression models, which are in fact special cases of probabilistic network models. 

Using the probabilistic network model, many hypothetical scenarios can be tested in order to... 

AL Delcher, S Kasif, HR Goldberg, WH Hsu. Protein secondary structure modelling with probabilistic networks. Intelligent Systems m Molecular Biology 1 109-117, 1993. 

MacKay s textbook [114] offers not only a comprehensive coverage of Shannon s theory of information but also probabilistic data modeling and the mathematical theory of neural networks. Artificial NN can be applied when problems appear with processing and analyzing the data, with their prediction and classification (data mining). The wide range of applications of NN also comprises optimization issues. The information-theoretic capabilities of some neural network algorithms are examined and neural networks are motivated as statistical models [114]. 

Niculescu, S.P., Kaiser, K.L.E., and Schultz, T.W., Modeling the toxicity of chemicals to Tetmhymena pyri-formis using molecular fragment descriptors and probabilistic networks, Arch. Environ. Contamination Toxicol., 39, 289-298, 2000. 

Price ND, Shmulevich I. Biochemical and statistical network models for systems biology. Curr. Opin. Biotechnol. 2007. Shmulevich I, et al. Probabilistic Boolean Networks a rule-based uncertainty model for gene regulatory networks. Bioinformatics. 2002 18 261-274... 

Friedman N. Inferring cellular networks using probabilistic graphical models. Science 2004 303 799-805. 98. 

Kaiser KLE, Niculescu SP, Schultz TW. Probabilistic neural network modeling for the toxicity of chemicals to Tetrahymena pyriformis with molecular fragment descriptors. SAR QSAR Environ Res 2002 13 57-67. 

Bayesian network inference is a very useful method to understand and visualize the relationship among various genetic variables of interest. In partieular, its Markov probabilistic inference is quite suitable in eapturing eausal relationships in gene network cascades. However, there still remain many problems to avoid excessive computational time to efficiently update the network structure and to improve accuracy and complexity in network structure inference. It is thus highly recommended that these Bayesian network models be validated both with known true-positive and simulated false-positive network variables and relationships. 

Friedman, F. 2004. Inferring Cellular Networks Using Probabilistic Graphical Models. Science 303 799-805. 

Most probabilistic models of complex systems assiune fiiU iadependence of their components. In particular, it means that the component lifetimes and repair times are independent random variables. Obviously, this assumption is made for the sake of computational simplicity. In reality, however, such independence rarely occurs. In order to make the considered network model more true-to-life, it is assumed that the functioning of Ci depends on the states of components located between Co andci in the following way if all these components are operable, then ei s time-to-failure is distributed according to Fi, otherwise (one of the components is failed) it is distributed according to Gi. It is also assumed that Fj > Gi, which conveys the idea that the components being under load are more failure prone, as in many real-life systems. Note that Gi = 0 if Ci disconnected from eo cannot fail. It is also assumed that ei functions independently of all components not located between ej and cq, and Cj s repair time is independent on the states of all other components. 

Vilar, S. Santana, L. Uriarte, E. Probabilistic neural network model for the in silico evaluation of anti-HIV activity and mechanism of action. J. Med. Chem. 2006, 49,1118-1124. 

Currently, the landslide hazard spatial prediction methods can be divided into qualitative methods and quantitative methods. As we all know qualitative forecasting method mainly depends on the subjective experience and the predicted accuracy of qualitative methods is lower than it of quantitative methods. So the qualitative methods have been gradually replaced by the quantitative methods. Quantitative models can be divided into statistic analysis models, deterministic models, probabilistic model, fuzzy information optimization processing and neurd network models. 

Chandrasekaran S, Price ND (2010) Probabilistic integrative modeling of genome-scale metabolic and regulatory networks in Escherichia coli and mycobacterium tuberculosis. Proc Natl Acad Sci USA 107 17845-17850... 

Niculescu, S. P., Kaiser, K. L. E., and Schuurmann, G. (1998) Influence of data preprocessing and kernel selection on probabilistic neural network modeling of the acute toxicity of chemicals to the fathead minnow and Vibrio fischeri bacteria. Water Qual. Res. J. Can. 33, 153-165. 

Zhuang, Y, J. Pan, L. Cai (2010). Minimizing energy consumption with probabilistic distance models in wireless sensor networks. In Proceedings of IEEE INFOCOM 2010, pp. 2453-2461. 

Bayesian statistics and Bayes Nets Bayesian statistics are founded on the Bayes theorem that can be used to calculate conditional probabilities. Bayesian statistics are often represented as Bayesian Belief Networks (BBNs). BBNs are directed acyclic graphs that represent probabilistic dependency models. What distinguishes the BBNs from other casual belief networks is the use of Bayesian calculus to determine the state probabilities of each node or variable from the predetermined conditional and prior probabilities (Krieg, 2001), meaning that the probabilities can be based on a person s belief of the likelihood of an event. 

ABSTRACT The overall aim of the ProGasNet (Probabilistic Gas Network Simulator) project is to develop the European gas transmission network probabilistic model capable to analyse reliability and risk aspects of the network under normal load and extreme crisis situations. The paper presents extension of the existing probabilistic Monte-Carlo based gas network model of 1-day to longer periods of up to 30 days when storage supply might be important to overcome crisis or network equipment failures. The study test case represents a country transmission network model and impact of storage supply under various crisis situations is presented and various network management options are discussed. 

A number of studies have already been completed by using probabilistic gas network model (Praks Kopustinskas, 2014) for 1-day duration. The present study case includes storage facility and 1-day model cannot capture the effect of reduced... 

An important consideration in many learning scenarios is the reliability of data and/or missing values. One learning method that is designed to reason in cases of uncertainty is that of Bayesian probabilistic networks. This method has been used to learn and model reaction mechanisms from physicochemical descriptions of instances of preclassified chemical reactions. ... 

BBNs are also sometimes called causal probabilistic networks, probabilistic cause-effect models or probabilistic influence diagrams. 

---