Cost-Sensitive Large margin Distribution Machine for classification of imbalanced data

Cheng, Fanyong; Zhang, Jing; Wen, Cuihong

doi:10.1016/j.patrec.2016.06.009

Cited by 75 publications

(13 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For important oversampling, number of nearest neighbors k is tuned from the candidate set {5, 8}. All the best parameters are selected via 5-fold cross validation from the training data at the first random data partition as CS-LDM [9] does. Because there are several parameters, in experiments, nested cross validation is adopted for model evaluation; i.e., 5-fold cross-validation is used for both validation set and testing set.…”

Section: Methodsmentioning

confidence: 99%

“…Till now, lots of methods have been proposed, which can be roughly divided into three main categories: cost sensitive-based, samplingbased and active learning-based methods. Among them, cost sensitive-based methods [8][9][10][11] are designed by introducing class-dependent or example-dependent cost to a traditional classification model. But it is hard to predefine the proper values for the cost variables.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Important sampling based active learning for imbalance classification

Wang

Liu

Cao

et al. 2020

Sci. China Inf. Sci.

View full text Add to dashboard Cite

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Important sampling based active learning for imbalance classification

Wang

Liu

Cao

et al. 2020

Sci. China Inf. Sci.

View full text Add to dashboard Cite

show abstract

“…Also, in [7,17], the prediction performance of the imbalanced data is improved by tuning a weight for each class. The margin distribution theory is used in [6] to design a balanced classifier by introducing a Cost Sensitive Large margin Distribution Machine (CS-LDM) to improve the detection rate of the minority class by using cost-sensitive margin mean and cost-sensitive penalty. Some approaches for the problem of imbalanced data apply the misclassifications as a loss function to update the distribution of the training data on successive boosting rounds [12,32,14].…”

Section: Related Workmentioning

confidence: 99%

Weighted Least Squares Twin Support Vector Machine with Fuzzy Rough Set Theory for Imbalanced Data Classification

Behmanesh,

Adibi,

Karshenas

2021

Preprint

View full text Add to dashboard Cite

Support vector machines (SVMs) are powerful supervised learning tools developed to solve classification problems. However, SVMs are likely to perform poorly in the classification of imbalanced data. The rough set theory presents a mathematical tool for inference in nondeterministic cases that provides methods for removing irrelevant information from data. In this work, we propose an approach that efficiently used fuzzy rough set theory in weighted least squares twin support vector machine called FRLSTSVM for classification of imbalanced data. The first innovation is introducing a new fuzzy rough set based under-sampling strategy to make the classifier robust in terms of the imbalanced data. For constructing the two proximal hyperplanes in FRLSTSVM, data points from the minority class remain unchanged while a subset of data points in the majority class are selected using a new method. In this model, we embed the weight biases in the LSTSVM formulations to overcome the bias phenomenon in the original twin SVM for the classification of imbalanced data. In order to determine these weights in this formulation, we introduce a new strategy that uses fuzzy rough set theory as the second innovation. Experimental results on the famous imbalanced datasets, compared to the related traditional SVM-based methods, demonstrate the superiority of the proposed

show abstract

“…The most common methods to address class imbalance are cost sensitive learning [9,11,44], algorithmic-level approaches [19,30,50], data resampling [6,13,36], and hybrid techniques composed of some combination of the aforementioned methods [25,28,41,49]. Cost sensitive learning assigns a diferent weight to misclassiication of minority samples than those from the majority class, and can operate at the resampling or algorithmic level.…”

Section: Introductionmentioning

confidence: 99%

Distance-based Probabilistic Data Augmentation for Synthetic Minority Oversampling

Goodman

Sarkani

Mazzuchi

2021

ACM/IMS Trans. Data Sci.

View full text Add to dashboard Cite

Class imbalance can adversely affect the performance of machine learning for prediction and classification. One approach to address the class imbalance problem is synthetic minority oversampling. Oversampling approaches can be broadly categorized as either being structural or statistical in nature. Structural approaches generally have the advantage of identifying and oversampling those minority data points that best facilitate class separation, while statistical approaches model the underlying distribution from which the minority samples can be drawn. In this paper, we formulate a distance-based approach which generates samples by both modeling the underlying minority class distribution and by geometrically considering those borderline samples entangled in the majority class. We demonstrate the efficacy of our approach operating on the Class-Imbalance data set from UCI by comparing its mean accuracy, AUC and F 1 -score performance against both statistical and structural synthetic minority oversampling methods.

show abstract

Cost-Sensitive Large margin Distribution Machine for classification of imbalanced data

Cited by 75 publications

References 8 publications

Important sampling based active learning for imbalance classification

Important sampling based active learning for imbalance classification

Weighted Least Squares Twin Support Vector Machine with Fuzzy Rough Set Theory for Imbalanced Data Classification

Distance-based Probabilistic Data Augmentation for Synthetic Minority Oversampling

Contact Info

Product

Resources

About