Results 1 
7 of
7
Utilitybased regression
 Proceedings of the Eleventh European Conference on Principles and Practice of Knowledge Discovery in Databases 4702:597
"... Abstract. Costsensitive learning is a key technique for addressing many real world data mining applications. Most existing research has been focused on classification problems. In this paper we propose a framework for evaluating regression models in applications with nonuniform costs and benefits ..."
Abstract

Cited by 8 (2 self)
 Add to MetaCart
(Show Context)
Abstract. Costsensitive learning is a key technique for addressing many real world data mining applications. Most existing research has been focused on classification problems. In this paper we propose a framework for evaluating regression models in applications with nonuniform costs and benefits across the domain of the continuous target variable. Namely, we describe two metrics for asserting the costs and benefits of the predictions of any model given a set of test cases. We illustrate the use of our metrics in the context of a specific type of applications where nonuniform costs are required: the prediction of rare extreme values of a continuous target variable. Our experiments provide clear evidence of the utility of the proposed framework for evaluating the merits of any model in this class of regression domains. 1
Detection and Prediction of Rare Events in Transaction Databases
, 2007
"... Rare events analysis is an area that includes methods for the detection and prediction of events, e.g. a network intrusion or an engine failure, that occur infrequently and have some impact to the system. There are various methods from the areas of statistics and data mining for that purpose. In thi ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
Rare events analysis is an area that includes methods for the detection and prediction of events, e.g. a network intrusion or an engine failure, that occur infrequently and have some impact to the system. There are various methods from the areas of statistics and data mining for that purpose. In this article we propose PREVENT, an algorithm which uses intertransactional patterns for the prediction of rare events in transaction databases. PREVENT is a general purpose intertransaction association rules mining algorithm that optimally fits the demands of rare event prediction. It requires only 1 scan on the original database and 2 over the transformed, which is considerably smaller and it is complete as it does not miss any patterns. We provide the mathematical formulation of the problem and experimental results that show PREVENT’s efficiency in terms of run time and effectiveness in terms of sensitivity and specificity.
A graphical analysis of costsensitive regression problems José HernándezOrallo
, 2012
"... Several efforts have been done to bring ROC analysis beyond (binary) classification, especially in regression. However, the mapping and possibilities of these proposals do not correspond to what we expect from the analysis of operating conditions, dominance, hybrid methods, etc. In this paper we pre ..."
Abstract
 Add to MetaCart
(Show Context)
Several efforts have been done to bring ROC analysis beyond (binary) classification, especially in regression. However, the mapping and possibilities of these proposals do not correspond to what we expect from the analysis of operating conditions, dominance, hybrid methods, etc. In this paper we present a new representation of regression models in the socalled regression ROC (RROC) space. The basic idea is to represent overestimation on the xaxis and underestimation on the yaxis. The curves are just drawn by adjusting a shift, a constant that is added (or subtracted) to the predictions, and plays a similar role as a threshold in classification. From here, we develop the notions of optimal operating condition, convexity, dominance, and explore several evaluation metrics that can be shown graphically, such as the area over the RROC curve (AOC). In particular, we show a novel and significant result, the AOC is equal to the error variance (multiplied by a factor which does not depend on the model). The derivation of RROC curves with nonconstant shifts and soft regression models, and the relation with cost plots is also discussed.
unknown title
, 2014
"... Soft (Gaussian CDE) regression models and loss functions ..."
(Show Context)
Contents
, 2014
"... Several efforts have been done to bring ROC analysis beyond (binary) classification, especially in regression. However, the mapping and possibilities of these proposals do not correspond to what we expect from the analysis of operating conditions, dominance, hybrid methods, etc. In this paper we pre ..."
Abstract
 Add to MetaCart
(Show Context)
Several efforts have been done to bring ROC analysis beyond (binary) classification, especially in regression. However, the mapping and possibilities of these proposals do not correspond to what we expect from the analysis of operating conditions, dominance, hybrid methods, etc. In this paper we present a new representation of regression models in the socalled regression ROC (RROC) space. The basic idea is to represent overestimation on the xaxis and underestimation on the yaxis. The curves are just drawn by adjusting a shift, a constant that is added (or subtracted) to the predictions, and plays a similar role as a threshold in classification. From here, we develop the notions of optimal operating condition, convexity, dominance, and explore several evaluation metrics that can be shown graphically, such as the area over the RROC curve (AOC). In particular, we show a novel and significant result, the AOC is equal to the error variance (multiplied by a factor which does not depend on the model). The derivation of RROC curves with nonconstant shifts and soft regression models, and the relation with cost plots is also discussed.