Inference bayesian network for multi-topographic neural network communication: a case study in documentary data

Shehabi, Shadi Al; Lamirel, Jean–Charles

doi:10.1109/ictta.2004.1307816

Cited by 4 publications

(20 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…An association rule is an expression A → B where A and B are conjunctions of properties. It means that if an individual data possesses all the properties of A then he necessarily possesses all the properties of B as regard to the studied dataset 1 . The support supp(A∪B) of the rule is equivalent to the number of individuals of the verifying both properties A and B, and the confidence conf(A∪B) is given by: conf(A∪B) = supp(A∪B)/supp(A).…”

Section: The Symbolic Model and Association Rules Extractionmentioning

confidence: 99%

“…The conservation of an overall view of the analysis is achieved through the use of a communication mechanism between the maps, which is itself based on Bayesian inference [1]. The advantage of the multi-viewpoint analysis provided by MultiSOM as compared to the global analysis provided by SOM [11][12] has been clearly demonstrated for precise mining tasks like patent analysis [19].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

New Fundamental Technologies in Data Mining

Funatsu¹

2011

View full text Add to dashboard Cite

Prof. Kimito Funatsu was born on December 3, 1955 in Japan. He received his Doctor degree (Dr. Sci.; physical organic chemistry) from Kyushu University which is one of the oldest imperial universities (1983) and joined Prof. Shin-ichi Sasaki's group, Toyohashi University of Technology (1984) as research-associate. During researching in Prof. Sasaki's group he worked on chemoinformatics field, i.e., Structure elucidation system CHEMICS, organic synthesis design system AIPHOS, molecular design. In 1990 he took the post of associate-professor. In 2004 he moved to The University of Tokyo as full Professor to continue the same research work. Recently he is trying material design and soft sensor. He is a chair of division of chemical information and computer sciences, Chemical Society of Japan.Dr. Kiyoshi Hasegawa was born on August 18, 1963 in Kanazawa, Japan. He received his Ph.D. from Toyohashi University of Technology (1995) and worked at Biostructure group in Chugai Pharmaceutical Company in Japan. While working in Chugai, he spent three months at Swiss Basel Research Center in Roche for developing new de novo design program (2003). He is now principal scientist in his group and is involved in many SBDD projects. He is also the Roche data mining working group member and the secretary general of Japan Chemical Information Science. His interests are de novo design, Chemogenomics, data mining and web application.

show abstract

Section: The Symbolic Model and Association Rules Extractionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

New Fundamental Technologies in Data Mining

Funatsu¹

2011

View full text Add to dashboard Cite

show abstract

“…clusters) T k which inherited of the activity (evidence Q) transmitted by their associated data or descriptor nodes. This computation can be carried out efficiently because of the specific Bayesian inference network topology that is associated to the set of models by the MVDA paradigm [1]. Hence, it is possible to compute the probability P(ac m t|t k ,Q) for an activity of modality act m on the model node t k which is inherited from activities generated on the source model.…”

Section: The Mvda Modelmentioning

confidence: 99%

“…An association rule is an expression A → B where A and B are conjunctions of properties. It means that if an individual data possesses all the properties of A then he necessarily www.intechopen.com possesses all the properties of B as regard to the studied dataset 1 . The support supp(A∪B) of the rule is equivalent to the number of individuals of the verifying both properties A and B, and the confidence conf(A∪B) is given by: conf(A∪B) = supp(A∪B)/supp(A).…”

Section: The Symbolic Model and Association Rules Extractionmentioning

confidence: 99%

“…The conservation of an overall view of the analysis is achieved through the use of a communication mechanism between the maps, which is itself based on Bayesian inference [1]. The advantage of the multi-viewpoint analysis provided by MultiSOM as compared to the global analysis provided by SOM [11] [12] has been clearly demonstrated for precise mining tasks like patent analysis [19].…”

mentioning

confidence: 99%

See 1 more Smart Citation