Academic Journal
Modelling count data using the logratio-normal-multinomial distribution
| Title: | Modelling count data using the logratio-normal-multinomial distribution |
|---|---|
| Authors: | Comas Cufí, Marc, Martín-Fernández, Josep Antoni, Mateu-Figueras, Glòria, Palarea-Albaladejo, Javier |
| Source: | SORT-Statistics and Operations Research Transactions; Vol. 44, Núm. 1 (2020): ; 99-126 SORT-Statistics and Operations Research Transactions; Vol 44, No 1 (2020): January-June; 99-126 oai:raco.cat:article/371184 Repositori Institucional de la Universitat Rovira i Virgili Universitat Rovira i virgili (URV) UPCommons. Portal del coneixement obert de la UPC Universitat Politècnica de Catalunya (UPC) Dipòsit Digital de Documents de la UAB Universitat Autònoma de Barcelona |
| Publisher Information: | Universitat Rovira i Virgili, 2020. |
| Publication Year: | 2020 |
| Subject Terms: | Estadística matemàtica, Simplex, Estadística matemàtica--Aplicacions, Compound probability distribution, 62 Statistics::62P Applications [Classificació AMS], Classificació AMS::62 Statistics::62F Parametric inference, Monte carlo method, Dirichlet multinomial, count data, Classificació AMS::62 Statistics::62P Applications, compound probability distribution, Monte Carlo method, logratio coordinates, Àrees temàtiques de la UPC::Matemàtiques i estadística::Estadística matemàtica, Logratio coordinates, Matemàtiques i estadística::Estadística matemàtica [Àrees temàtiques de la UPC], 62 Statistics::62F Parametric inference [Classificació AMS], simplex, Count data |
| Description: | The logratio-normal-multinomial distribution is a count data model resulting from compounding a multinomial distribution for the counts with a multivariate logratio-normal distribution for the multinomial event probabilities. However, the logratio-normal-multinomial probability mass function does not admit a closed form expression and, consequently, numerical approximation is required for parameter estimation. In this work, different estimation approaches are introduced and evaluated. We concluded that estimation based on a quasi-Monte Carlo Expectation-Maximisation algorithm provides the best overall results. Building on this, the performances of the Dirichlet-multinomial and logratio-normal-multinomial models are compared through a number of examples using simulated and real count data. |
| Document Type: | Article |
| File Description: | application/pdf |
| DOI: | 10.2436/20.8080.02.96 |
| Access URL: | http://hdl.handle.net/20.500.11797/RP4899 https://ddd.uab.cat/record/225688 https://dialnet.unirioja.es/servlet/articulo?codigo=7537020 https://ddd.uab.cat/pub/sort/sort_a2020v44n1/sort_a2020v44n1p99.pdf https://ddd.uab.cat/record/225688?ln=ca https://hdl.handle.net/2117/362096 https://doi.org/10.2436/20.8080.02.96 |
| Rights: | CC BY NC ND |
| Accession Number: | edsair.dedup.wf.002..6101988af52b6d38c7dad1ae4842a974 |
| Database: | OpenAIRE |
| DOI: | 10.2436/20.8080.02.96 |
|---|