Articolul precedent |
Articolul urmator |
![]() |
![]() ![]() |
![]() BEJAN, Andrei. Bayesian Experimental Design for Network Loss Tomography. In: Conference of Mathematical Society of the Republic of Moldova, 19-23 august 2014, Chișinău. Chișinău: "VALINEX" SRL, 2014, 3, pp. 310-313. ISBN 978-9975-68-244-2. |
EXPORT metadate: Google Scholar Crossref CERIF DataCite Dublin Core |
Conference of Mathematical Society of the Republic of Moldova 3, 2014 |
||||||
Conferința "Conference of Mathematical Society of the Republic of Moldova" Chișinău, Moldova, 19-23 august 2014 | ||||||
|
||||||
Pag. 310-313 | ||||||
|
||||||
![]() |
||||||
Rezumat | ||||||
We consider a sequential experimental design problem for network loss tomography with a random choice of paths. We apply a Bayesian approach and Kullback-Leibler divergence to maximise the informational gain obtained at each step of the network probing experiment and show that the choice of paths is, in fact, deterministic. We discuss practical aspects of this result. |
||||||
Cuvinte-cheie path probing, statistical network loss tomography, Bayesian experimental design, Kullback{Leibler divergence |
||||||
|
Google Scholar Export
<meta name="citation_title" content="Bayesian Experimental Design for Network Loss Tomography"> <meta name="citation_author" content="Bejan Andrei"> <meta name="citation_publication_date" content="2014"> <meta name="citation_collection_title" content="Conference of Mathematical Society of the Republic of Moldova"> <meta name="citation_volume" content="3"> <meta name="citation_firstpage" content="310"> <meta name="citation_lastpage" content="313"> <meta name="citation_pdf_url" content="https://ibn.idsi.md/sites/default/files/imag_file/310_313_Bayesian%20Experimental%20Design%20for%20Network%20Loss%20Tomography.pdf">