Browsing by Author "Tjur, Tue"
Now showing items 1-5 of 5
-
Kronborg, Dorte; Tjur, Tue (Frederiksberg, 1999)[More information][Less information]
Abstract: The scenario considered is that of a credit association, a bank or an- other nancial institution which, on the basis of information about a new potential customer and historical data on many other customers, has to decide whether or not to give that customer a certain loan. We discuss three popular techniques: logistic regression, discriminant analysis and neural networks. We shall argue strongly in favour of the logistic regression. Discriminant analysis can be used, and for reasons that can be explained mathematically it will often result in approximately the same conclusions as a logistic regression. But the statistical assumptions are not appropriate in most cases, and the results given are not as directly interpretable as those of logistic re- gression. Neural network techniques, in their simplest form, su er from the lack of statistical standard methods for veri cation of the model and tests for removal of covariates. This problem disappears to some extend when the neural networks are reformulated as proper statistical models, based on the type of functions that are considered in neural networks. But this results in a somewhat specialized class of non{linear regression models, which may be useful in situations where local peculiarities of the response function are in focus, but certainly not when the overall | usually monotone | e ect of many more or less confounded covariates is the issue. We discuss, within the logistic regression framework, the handling of phenomena such as time trends and corruption of the historical data due to shifts of policy, censor- ing and/or interventions in highrisk customers' economy. Finally, we illustrate and support the theoretical considerations by a case study concerning mortgage loans in a Danish credit associatio URI: http://hdl.handle.net/10398/8131 Files in this item: 1
x644964528.pdf (343.3Kb) -
[More information][Less information]
-
Birch, Kristina; Olsen, Jørgen Kai; Tjur, Tue (København, 2005)[More information][Less information]
Abstract: On the background of a data set of weekly sales and prices for three brands of coffee, this paper discusses various regression models and their relation to the multiplicative competitive-interaction model (the MCI model, see Cooper 1988, 1993) for market-shares. Emphasis is put on the interpretation of the parameters in relation to models for the total sales based on discrete choice models. Key words and phrases. MCI model, discrete choice model, market-shares, price elasitcity, regression model. URI: http://hdl.handle.net/10398/6735 Files in this item: 1
stat-pp-2005-1.pdf (274.7Kb) -
Hougaard, Jens Leth; Tjur, Tue; Østerdal, Lars Peter (København, 2004)[More information][Less information]
Abstract: Discrete choice experiments are widely used in relation to health care. A stream of recent literature therefore aims at testing the validity of the underlying preference axioms of completeness and transitivity, and detecting other preference phenomena such as unstability, learn- ing/tiredness effects, ordering effects, dominance, etc. Unfortunately there seems to be some confusion about what is actually being tested, and the link between the statistical tests performed and the relevant underlying model of respondent behaviour has not been explored in this literature. The present paper tries to clarify the notions involved and discuss what can be tested in a general frequency of choice frame- work and more specifically in a random utility model. URI: http://hdl.handle.net/10398/6731 Files in this item: 1
04-1.pdf (151.6Kb) -
Tjur, Tue (København, 2003)[More information][Less information]
Now showing items 1-5 of 5