Paola Cerchiello | Università degli studi di Pavia (original) (raw)

Uploads

Papers by Paola Cerchiello

Communications in Statistics - Theory and Methods, 2014

ABSTRACT In this contribution we aim at improving ordinal variable selection in the context of ca... more ABSTRACT In this contribution we aim at improving ordinal variable selection in the context of causal models for credit risk estimation. In this regard, we propose an approach that provides a formal inferential tool to compare the explanatory power of each covariate and, therefore, to select an effective model for classification purposes. Our proposed model is Bayesian nonparametric thus keeps the amount of model specification to a minimum. We consider the case in which information from the covariates is at the ordinal level. A noticeable instance of this regards the situation in which ordinal variables result from rankings of companies that are to be evaluated according to different macro and micro economic aspects, leading to ordinal covariates that correspond to various ratings, that entail different magnitudes of the probability of default. For each given covariate, we suggest to partition the statistical units in as many groups as the number of observed levels of the covariate. We then assume individual defaults to be homogeneous within each group and heterogeneous across groups. Our aim is to compare and, therefore select, the partition structures resulting from the consideration of different explanatory covariates. The metric we choose for variable comparison is the calculation of the posterior probability of each partition. The application of our proposal to a European credit risk database shows that it performs well, leading to a coherent and clear method for variable averaging of the estimated default probabilities.

The Journal of Operational Risk, 2012

Bancaria, 2012

To estimate the probability of default of companies and the correlated rating classes, it is nece... more To estimate the probability of default of companies and the correlated rating classes, it is necessary to use efficiently the information contained in different databases. In this respect, we propose a novel approach, based on the recursive usage of Bayes theorem, that can be very helpful in integrating default estimates obtained from different sets of covariates. The application of our proposal to an Italian credit risk database shows that it performs quite efficiently, allowing to predict for each company the probability of default by averaging the ...

Advances in Latent Variables Methods Models and Applications, Jun 13, 2013

Abstract This contribution deals either with the presentation or the analysis of a “Workplace Ris... more Abstract This contribution deals either with the presentation or the analysis of a “Workplace Risk Perception Questionnaire” focused on the identification and quantification of risk perception with particularly regard to work injuries, safety training and safety compliance. The main purpose of this paper is to inquire into the injury risk and the related perception by using statistical models in order to get knowledge about the work categories more exposed to injuries risk. We got some important results that witness the importance of the level of ...

Expert Systems with Applications, 2015

SUMMARY The aim of this paper is to present a new proposal for the classiflcation of academic ins... more SUMMARY The aim of this paper is to present a new proposal for the classiflcation of academic institutions in terms of quality of teaching. Our methodological proposal borrows con- cepts from operational risk, such as scorecard models, employed to assess University performances, on the basis of the perceived quality. We propose to summarize opinion data using new non parametric indexes able to exploit e-ciently the ordinal nature of the analyzed variables. Through the application of such indexes we obtain a complete ranking of University courses. Empirical evidence is flnally given on the basis of real data from the University of Pavia. In this paper we present models for the evaluation of the quality of Academic teaching. The objective of the proposed methodology is to evaluate and, there- fore, to improve the quality of university institutions. The quality of university teaching depends mainly on the performances of its students who thereby rep- resent the main actors of the analy...

Studies in Classification, Data Analysis, and Knowledge Organization, 2009

This contribution is deemed in the view of the authors, as a methodological proposal in order to ... more This contribution is deemed in the view of the authors, as a methodological proposal in order to employ the well know fuzzy approach in a context of operational risk management. Even though the available data can not be considered native fuzzy, we show that modelling them according to fuzzy intervals is useful from two point of view: it allows to take into account and to exploit more information and, on the other hand, either unsupervised or supervised models applied to this kind of data present comparatively good performance. The paper ...

Studies in Classification, Data Analysis, and Knowledge Organization, 2013

Open Journal of Statistics, 2012

In this contribution we aim at improving ordinal variable selection in the context of causal mode... more In this contribution we aim at improving ordinal variable selection in the context of causal models. In this regard, we propose an approach that provides a formal inferential tool to compare the explanatory power of each covariate, and, therefore, to select an effective model for classification purposes. Our proposed model is Bayesian nonparametric, and, thus, keeps the amount of model specification to a minimum. We consider the case in which information from the covariates is at the ordinal level. A noticeable instance of this regards ...