Infinitely Imbalanced Logistic Regression Export

@article{citeulike:4207853, abstract = {In binary classification problems it is common for the two classes to be imbalanced: one case is very rare compared to the other. In this paper we consider the infinitely imbalanced case where one class has a finite sample size and the other class's sample size grows without bound. For logistic regression, the infinitely imbalanced case often has a useful solution. Under mild conditions, the intercept diverges as expected, but the rest of the coefficient vector approaches a non trivial and useful limit. That limit can be expressed in terms of exponential tilting and is the minimum of a convex objective function. The limiting form of logistic regression suggests a computational shortcut for fraud detection problems.}, author = {Owen, Art B.}, citeulike-article-id = {4207853}, citeulike-linkout-0 = {http://www.jmlr.org/papers/volume8/owen07a/owen07a.pdf}, journal = {Journal of Machine Learning Research}, keywords = {classification, imbalance, logistic, regression}, month = {April}, pages = {761--773}, posted-at = {2009-03-22 23:05:30}, priority = {3}, title = {Infinitely Imbalanced Logistic Regression}, url = {http://www.jmlr.org/papers/volume8/owen07a/owen07a.pdf}, volume = {8}, year = {2007} }

TY - JOUR ID - citeulike:4207853 L3 - citeulike-article-id:4207853 N2 - In binary classification problems it is common for the two classes to be imbalanced: one case is very rare compared to the other. In this paper we consider the infinitely imbalanced case where one class has a finite sample size and the other class's sample size grows without bound. For logistic regression, the infinitely imbalanced case often has a useful solution. Under mild conditions, the intercept diverges as expected, but the rest of the coefficient vector approaches a non trivial and useful limit. That limit can be expressed in terms of exponential tilting and is the minimum of a convex objective function. The limiting form of logistic regression suggests a computational shortcut for fraud detection problems. JF - Journal of Machine Learning Research EP - 773 TI - Infinitely Imbalanced Logistic Regression VL - 8 SP - 761 KW - classification KW - imbalance KW - logistic KW - regression AU - Owen, Art PY - 2007/04// UR - http://www.jmlr.org/papers/volume8/owen07a/owen07a.pdf ER -