Suchergebnisse

In this paper we propose a framework to generate artificial data that can be used to simulate credit risk scenarios. Artificial data is useful in the credit scoring domain for two reasons. Firstly, the use of artificial data allows for the introduction and control of variability that can realistically be expected to occur, but has yet to materialise in practice. The ability to control parameters allows for a thorough exploration of the performance of classification models under different conditions. Secondly, due to non-disclosure agreements and commercial sensitivities, obtaining real credit scoring data is a problematic and time consuming task. By the provision of publicly available artificial data, credit scoring is opened to the wider data mining community. This in turn could help enable greater participation, promote replicable experimental findings, and give rise to solution proposals to outstanding credit scoring problems. To ensure that our framework is sufficiently grounded in reality, data distributions are generated using a troika of sources: demographic information from the Central Statistics Office, Ireland; housing statistics published by the Irish Government Department of the Environment, Heritage and Local Government; and a profile of loan defaulters developed using a recent report published by a credit rating agency. By engaging with a credit scoring expert we select characteristics that are typical of most application scorecard models including, amongst others: age, income, loan value, and occupation. Through user controlled settings the conditional prior probabilities of the characteristics can be adjusted over time to simulate differing scenarios. In order to assign class labels to the generated data a credit risk score is estimated based on the non-linear interactions between various characteristics. Based on the desired number of defaulters a cut-off score is placed on this monotonic ordering of credit scores to distinguish between those likely to repay and those likely to default on their financial obligation. The classification complexity is controlled by adding user-defined random Gaussian noise. After discussing the desirable characteristics of artificial data we describe a pseudo-random data generator for credit scoring and provide illustrations on how the framework can be used to generate population drift.

AbstractSummary statistics and data visualizations are often used to explore data and draw preliminary conclusions. Although valuable, these tools do not always reveal the underlying patterns and trends in the data and can sometimes be misleading. We describe an approach for teaching the need for more advanced statistical analysis using multiple linear regression. Our approach is based on using a method we developed for generating alternative multivariate data sets where all the variables (both independent and dependent) have the same summary statistics. However, we can deliberately change the statistical significance of one (or more) of the independent variables in the regression to illustrate why it is important to go beyond simple descriptive measures and examine inferential statistics on the inherent relationships in the data. Implementation of this methodology is provided in the R statistical programming language and an add‐in for Excel spreadsheets.

This article argues that bottom–up, people-centered research which uses ethnographic and everyday approaches is crucial but underutilized in research on identity politics in Eastern Europe. In order to understand what concepts such as ethnicity and citizenship mean in the context of people's everyday lives, it is vital to understand whether taken-for-granted political concepts are appropriate and the make-up of data such as census data. The article first introduces the methods of political ethnography and bottom–up interviews by discussing how they can be applied and their value within political science. The paper uses data gathered from interviews in Moldova and Crimea (when it was still a de jure and de facto part of Ukraine) to demonstrate the value of this approach. It shows how interview data can add significantly to the understanding of kin-state relations within political science by adding a richness of context and a bottom–up perspective that quantitative and elite-level interviews fail to provide. Lastly, the paper draws on experiences gained from research design to discuss how bottom–up research in political science can be conducted rigorously. The article argues that this approach can deepen the understanding of identity politics and kin-state relations or, more broadly, important post-communist questions such as democratization and Europeanization.

Ausgehend von dem Argument, dass die Rolle informeller Gespräche in qualitativen Methoden der Sozial- und Bildungsforschung umstritten, aber auch relativ vernachlässigt ist, stellen wir in diesem Artikel dar, wie diese Gespräche unsere Forschungsansätze und -praxis beeinflusst haben. Wir verwenden ein Beispiel aus einer eigenen Studie, um ihre nuancierte und spezifische Natur hervorzuheben und eine Reihe von sich ergebenden ethischen und methodischen Fragen aufzuwerfen und zu hinterfragen. Wir betrachten informelle Gespräche als Möglichkeit, Daten mit "Kontext" und "Authentizität" zu versehen und andernfalls verpasste Möglichkeiten zur Erweiterung und Anreicherung von Daten zu erschließen. Wir berücksichtigen auch die Rolle von Ethikkommissionen und Ethikrichtlinien sowie die praktischen Auswirkungen und Konsequenzen, die diese für Forscher_innen haben, wenn sie während ihrer Feldarbeit informelle Gespräche führen.

International audience ; In this paper, we present a platform we have designed and built in order to generate data and scenario traces that can serve as inputs and references when evaluating our algorithms for detecting cyber security intrusions. Our context is related to civilian and military ships and our research is performed within a strong partnership with industry and academy. As obtaining actual and accurate data sets are not straightforward, we have chosen do generate realistic datasets with a platform we have designed and built.

International audience ; In this paper, we present a platform we have designed and built in order to generate data and scenario traces that can serve as inputs and references when evaluating our algorithms for detecting cyber security intrusions. Our context is related to civilian and military ships and our research is performed within a strong partnership with industry and academy. As obtaining actual and accurate data sets are not straightforward, we have chosen do generate realistic datasets with a platform we have designed and built.

International audience ; In this paper, we present a platform we have designed and built in order to generate data and scenario traces that can serve as inputs and references when evaluating our algorithms for detecting cyber security intrusions. Our context is related to civilian and military ships and our research is performed within a strong partnership with industry and academy. As obtaining actual and accurate data sets are not straightforward, we have chosen do generate realistic datasets with a platform we have designed and built.