In statistics AU80: The URL http://en.wikipedia.org/wiki/Statistics has been redirected to https://en.wikipedia.org/wiki/Statistics. Please verify the URL. , logistic regression, or logit regression, or logit model is a type of probabilistic statistical classification AU81: The URL http://en.wikipedia.org/wiki/Statistical_classification has been redirected to https://en.wikipedia.org/wiki/Statistical_classification. Please verify the URL. model. It is also used to predict a binary response from a binary predictor AU82: The URL http://en.wikipedia.org/wiki/Binary_classification has been redirected to https://en.wikipedia.org/wiki/Binary_classification. Please verify the URL. , used for predicting the outcome of a categorical AU83: The URL http://en.wikipedia.org/wiki/Categorical_variable has been redirected to https://en.wikipedia.org/wiki/Categorical_variable. Please verify the URL. dependent variable AU84: The URL http://en.wikipedia.org/wiki/Dependent_and_independent_variables has been redirected to https://en.wikipedia.org/wiki/Dependent_and_independent_variables. Please verify the URL. (i.e., a class label) based on one or more predictor variables (features). That is, it is used in estimating the parameters of aqualitative response model AU85: The URL http://en.wikipedia.org/wiki/Qualitative_response_models has been redirected to https://en.wikipedia.org/wiki/Qualitative_response_models. Please verify the URL. . The probabilities describing the possible outcomes of a single trial are modeled, as a function of the explanatory (predictor) variables, using a logistic function AU86: The URL http://en.wikipedia.org/wiki/Logistic_function has been redirected to https://en.wikipedia.org/wiki/Logistic_function. Please verify the URL. . Frequently (and hereafter in this article) “logistic regression” is used to refer specifically to the problem in which the dependent variable is binary AU87: The URL http://en.wikipedia.org/wiki/Binary_variable has been redirected to https://en.wikipedia.org/wiki/Binary_variable. Please verify the URL. —that is, the number of available categories is two—while problems with more than two categories are referred to as multinomial logistic regression AU88: The URL http://en.wikipedia.org/wiki/Multinomial_logistic_regression has been redirected to https://en.wikipedia.org/wiki/Multinomial_logistic_regression. Please verify the URL. or, if the multiple categories are ordered as ordered logistic regression AU89: The URL http://en.wikipedia.org/wiki/Ordered_logistic_regression has been redirected to https://en.wikipedia.org/wiki/Ordered_logistic_regression. Please verify the URL. . Logistic regression measures the relationship between the categorical dependent variable and one or more independent variables, which are usually (but not necessarily) continuous, by using probability scores as the predicted values of the dependent variable.Thus, it treats the same set of problems as does probit regression AU90: The URL http://en.wikipedia.org/wiki/Probit_regression has been redirected to https://en.wikipedia.org/wiki/Probit_regression. Please verify the URL. using similar techniques; the first assumes a logistic function AU91: The URL http://en.wikipedia.org/wiki/Logistic_function has been redirected to https://en.wikipedia.org/wiki/Logistic_function. Please verify the URL. and the second a standard normal distribution AU92: The URL http://en.wikipedia.org/wiki/Normal_distribution has been redirected to https://en.wikipedia.org/wiki/Normal_distribution. Please verify the URL. function. Logistic regression can be seen as a special case of generalized linear model AU93: The URL http://en.wikipedia.org/wiki/Generalized_linear_model has been redirected to https://en.wikipedia.org/wiki/Generalized_linear_model. Please verify the URL. and thus analogous to linear regression AU94: The URL http://en.wikipedia.org/wiki/Linear_regression has been redirected to https://en.wikipedia.org/wiki/Linear_regression. Please verify the URL. . The model of logistic regression, however, is based on quite different assumptions (about the relationship between dependent and independent variables) from those of linear regression. In particular the key differences of these two models can be seen in the following two features of logistic regression. First, the conditional mean p(y/x) follows a Bernoulli distribution AU95: The URL http://en.wikipedia.org/wiki/Bernoulli_distribution has been redirected to https://en.wikipedia.org/wiki/Bernoulli_distribution. Please verify the URL. rather than a Gaussian distribution AU96: The URL http://en.wikipedia.org/wiki/Gaussian_distribution has been redirected to https://en.wikipedia.org/wiki/Gaussian_distribution. Please verify the URL. , because logistic regression is a classifier. Second, the linear combination of the inputs is restricted to [0,1] through the logistic distribution function AU97: The URL http://en.wikipedia.org/wiki/Logistic_function has been redirected to https://en.wikipedia.org/wiki/Logistic_function. Please verify the URL. because logistic regression predicts the probability of the instance being positive.
Published in Chapter:
KLR Approach as an Early Warning Indicator of Turkish Currency and Banking Crisis in 2000 and 2001
Filiz Eryılmaz (Uludağ University, Turkey)
Copyright: © 2016
|Pages: 18
DOI: 10.4018/978-1-4666-9484-2.ch011
Abstract
International organizations as private sector institutions started to develop Early Warning System [EWS] models aiming to anticipate whether and when individual countries can collide with a financial crisis. EWS models can be made most useful to help sustain global growth and maintain financial stability, especially in light of the lessons learned from the current and past crises. This paper proposes Early Warning Systems (EWS) for Turkish Currency and Banking Crisis in 2000 and 2001. To that end “KLR model” or “signaling window” approach developed by Kaminski, Lorezondo and Reinhart (1998) is testified in the empirical part of this research and applied to a sample of Turkey macroeconomic data for the 1998-2003 monthly periods.