170,589 research outputs found
Mystery shopping at greek banks : a Bayesian network analysis
No full textThe banking industry is highly competitive, and customer satisfaction plays an essential rule. Hence, methods to evaluate customer satisfaction are im-portant for bank managers. In this work, two instruments for customer satisfaction analysis are combined: Mystery shopping methodology and Bayesian Networks.
Mystery shoppers are used to survey and monitor the quality of customer service and to identify areas requiring enhancement. After each visit they complete a report prepared in advance on their service experience. Bayesian Networks are then used to provide a pictorial representation of the dependence structure between the variables of interest, and are used to study the effect of different improvement strategies. We present a real data analysis concerning customer satisfaction in Greek banks
Default probability estimation: bayesian Pair Copula model
No full textIn this paper we present a novel Bayesian methodology for default prob-
ability estimation based on multivariate contingent claim analysis and pair copula
theory. In order to compute the default probability of a firm, we use balance sheet
data as a proxy of the equity value. A pair copula approach is applied to obtain the
firm pricing function, and Monte Carlo simulations are then used to calculate the
distribution of the default probability. The methodology will be illustrated through
an application to real dat
MCMC Model determination for Discrete Graphical Models
No full textAbstract: In this paper we compare two alternative MCMC samplers for the Bayesian analysis of discrete
graphical models; we present both a hierarchical and a nonhierarchical version of them. We first consider
the MC3 algorithm by Madigan and York (1995) for which we propose an extension that allows for a
hierarchical prior on the cell counts. We then describe a novel methodology based on a reversible jump
sampler. As a prior distribution we assign, for each given graph, a hyper-Dirichlet distribution on the
matrix of cell probabilities. Two applications to real data are presented
Spanning trees and identifiability of a single-factor model
No full textThe aim of this paper is to propose conditions for exploring the class
of identifiable Gaussian models with one latent variable. In particular, we focus
attention on the topological structure of the complementary graph of the residuals.
These conditions are mainly based on the presence of odd cycles and bridge edges
in the complementary graph. We propose to use the spanning tree representation
of the graph and the associated matrix of fundamental cycles. In this way it is
possible to obtain an algorithm able to establish in advance whether modifying
the graph corresponding to an identifiable model, the resulting graph still denotes
identifiability
A semi-bayesian approach for the analysis of scale effects in ordinal regression models
No full textIn this paper we propose a semi-Bayesian approach for the analysis of
categorical data with an ordered outcome when a scaling component is considered. A
recursive partitioning method yielding two trees –one for the location and one for the
scaling– is used for selecting covariates, then a Bayesian approach for model estima-
tion is implemented and an MCMC sampler is used to obtain posterior estimates. An
analysis on risk perception concerning Covid-19 pandemic is carried out to assess the
performance of the metho
Marginal effects for comparing groups in regression models for ordinal outcome when uncertainty is present
No full textThis contribution deals with effect measures for covariates in ordinal
data models to address the interpretation of the results on the extreme categories of
the scales. It provides a simpler interpretation than model parameters both in standard
cumulative models with proportional odds assumption and in the recent extension
of the CUP models, the mixture models to account for uncertainty in the process of
selection of the score. Visualization tools for the effect of covariates are proposed and
the measure of relative size and marginal effects based on rates of change are evaluated
by use of a case study
- …
