Search results for " Bayesian"
showing 4 items of 124 documents
Elaborazione e validazione di modelli dell’impatto antropico e dei cambiamenti climatici sulle qualità dei suoli in Sicilia, con particolare attenzio…
Il declino della sostanza organica e l’erosione idrica sono due fattori di rischio di degrado del suolo fra i più diffusi in Sicilia. In che misura è possibile contrastarli modificando le attività antropiche ed in particolare quelle in ambito agricolo? Lo scopo della presente ricerca di dottorato è stato quello di mettere a punto dei modelli che permettessero di spiegare le relazioni esistenti fra le azioni dell’uomo in ambito agricolo con la sostanza organica e con l’erosione idrica nei suoli siciliani, tenendo conto anche del clima. Le due indagini sono state svolte separatamente, in parallelo, in aree agricole della Sicilia occidentale per la sostanza organica, sull’intero territorio sic…
Production and Efficiency in 21 Industrialized Countries from 1966 to 2007 to Analyse the Italy’s Decline: Application of a Stochastic Frontier Model…
2008
This paper uses a stochastic frontier production model to estimate the technical efficiency of 21 nations among the more industrialized countries in the world for the wide period from 1966 to 2007. A Cobb-Douglas with labour and capital as inputs, GDP as output and efficiency with exponential distribution appears to work good from a statistical point of view. This model is used to estimate in a bayesian approach the efficiency score of these 21 countries. The bayesian way, applied via Gibbs sampling, is here useful to avoid problems about the small number of observations. So, each efficiency score is obtained as mean of 10,000 draws. The results underline a general and progressive efficienc…
La falta de reproducibilitat de la investigació: L'estadística com a legitimació del resultat
2014
La investigació científica es legitima mitjançant la replicació dels seus resultats, però els esforços per replicar afirmacions enganyoses exhaureixen el finançament. Ens centrarem en una d’aquestes errades: els resultats de proves estadístiques que ofereixen falsos positius a causa de l’atzar. Els mètodes estadístics clàssics confien en un p-valor per a ponderar les proves enfront d’una hipòtesi nul·la, però les proves d’hipòtesis bayesianes ofereixen resultats més fàcils de comprendre, sempre que hom puga especificar distribucions a priori per a la hipòtesi alternativa. Descriurem noves proves, les UMPBT, tests bayesians que ofereixen una especificació per defecte de les alternatives a pr…
Características de los vecindarios y la distribución espacial de problemas sociales en la ciudad de Valencia
2018
The aim of this doctoral thesis is to explore the influence of neighborhood-level variables on the spatial and spatio-temporal distribution of different social problems in the city of Valencia. In Study 1, we present data on the development and validation of an observational instrument to assess neighborhood disorder. Results supported a three-factor model (physical disorder, social disorder and physical deterioration), and they showed good reliability and validity evidences. In Study 2, we assess the psychometric properties of a neighborhood disorder scale using Google Street View. Results supported a bifactorial solution with a general factor (general neighborhood disorder) and two specif…