Abstract We compare four conditional tests for Poisson data through a simulation study: the exact binomial test, its asymptotic approximation, a Markov Chain Monte Carlo test and the standard permutation test. Despite being non-parametric, we observe that permutation tests are as effective as the others. From a theoretical point of view we justify this result by observing that the orbits of permutations form a good partition of the conditional space.
Comparison of conditional tests on Poisson data / Crucinio, Francesca Romana; Fontana, Roberto. - ELETTRONICO. - (2017), pp. 333-338. ((Intervento presentato al convegno Statistics and Data Science: new challenges, new generations 28–30 June 2017 tenutosi a Firenze nel 28–30 June 2017.
Titolo: | Comparison of conditional tests on Poisson data | |
Autori: | ||
Data di pubblicazione: | 2017 | |
Abstract: | Abstract We compare four conditional tests for Poisson data through a simulation study: the exac...t binomial test, its asymptotic approximation, a Markov Chain Monte Carlo test and the standard permutation test. Despite being non-parametric, we observe that permutation tests are as effective as the others. From a theoretical point of view we justify this result by observing that the orbits of permutations form a good partition of the conditional space. | |
ISBN: | 978-88-6453-521-0 | |
Appare nelle tipologie: | 4.1 Contributo in Atti di convegno |
File in questo prodotto:
File | Descrizione | Tipologia | Licenza | |
---|---|---|---|---|
estratto da 3407_11724.pdf | Articolo | 2. Post-print / Author's Accepted Manuscript | ![]() | Visibile a tuttiVisualizza/Apri |
http://hdl.handle.net/11583/2675994