ON MULTI‐CLASS COST‐SENSITIVE LEARNING

By Zhi-Hua Zhou, Xu-Ying Liu. International Conference on Climate Informatics, 2006

358 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Civil and Structural Engineering, Electrical and Electronic Engineering

View Paper: https://www.semanticscholar.org/paper/f18fefc326bd59925bc4c65298dc974d38fe265c

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Civil and Structural Engineering

In Electrical and Electronic Engineering

Citations in the corpus

Analysing the classification of imbalanced data-sets with multiple classes: Binarization techniques and ad-hoc approaches (2013, methodology): “Next, in Section 3.2, we present a global cost-sensitive approach that re-weights the instances from each class according to their ratio [71].”
Learning from Imbalanced Data (2009, background): “Although this description would seem to imply that all between-class imbalances are innately binary (or two-class), we note that there are multiclass data in which imbalances exist between the various classes [7], [8], [9], [10], [11], [12].”
Ordinal hyperplanes ranker with cost sensitivities for age estimation (2011, methodology): “Note that applying the rescaling technique to multi-class data directly is not appropriate because it causes the inconsistency of costs problem [33].”
Multiclass Imbalance Problems: Analysis and Potential Solutions (2012, methodology): “To have a sufﬁcient number of classes for our study, we generate some artiﬁcial imbalanced data sets by using the method in [33].”
Learning to Segment the Tail (2020, background): “The former aims to re-balance the training samples across classes [16, 3, 11, 6]; while the latter focuses on assigning different weights to adjust the loss function [18, 40, 47, 7].”
Deep Imbalanced Learning for Face Recognition and Attribute Prediction (2018, background): “mainly divided into two groups: data re-sampling [5], [6], [7], [9], [10], [11], [15] and cost-sensitive learning [8], [12], [13], [14], [16].”
Multi-instance multi-label learning (2008, methodology): “One of the most popular and effective methods is rescaling [87], which can be incorporated into our framework easily.”
On multi-class classification through the minimization of the confusion matrix norm (2013, background): “As such it can be seen as a common base for most imbalanced classes learning problems, since it encompasses both resampling and penalization, as advocated in Zhou and Liu (2010).”
Look at the First Sentence: Position Bias in Question Answering (2020, methodology): “Using prior distributions in machine learning has a long history such as using class frequency in the class imbalance problem (Domingos, 1999; Japkowicz and Stephen, 2002; Zhou and Liu, 2006; Huang et al., 2016).”
Learning from imbalanced data: open challenges and future directions (2016, background): “The most popular branch is cost-sensitive approaches [67].”
Rethinking Class-Balanced Methods for Long-Tailed Visual Recognition From a Domain Adaptation Perspective (2020, background): “Alter-natively, one can derive these weights from the cost of mis-classifying an example of one class to another [13, 62].”
Precision Medicine. (2019, background): “…∈ D E Y 1Asign Y d X < 0 π A; X , hence, it can be seen that d𝒟 opt minimizes a cost-sensitive classification problem (Elkan 2001; Zadrozny et al 2003; Zhou and Liu 2006; Pires et al 2013) with cost function |Y|/π(A; X), class label A sign(Y), and inputX ∈ 𝒳, over the set of classifiers 𝒟…”