Robust support vector machines based on the rescaled hinge loss function

Xu, Guibiao; Cao, Zheng; Hu, Bao-Gang; Prı́ncipe, José C.

doi:10.1016/j.patcog.2016.09.045

Cited by 94 publications

(41 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Note how the shape of these balls influences the orientation of the decision boundaries, i.e., how different regularizers optimally counter specific kinds of adversarial noise. are intrinsically less sensitive to outlying training samples, e.g., via bounded losses or kernel functions) [25,117,[119][120][121][122].…”

Section: Security-by-design Defenses Against White-box Attacksmentioning

confidence: 99%

Wild patterns: Ten years after the rise of adversarial machine learning

Biggio

Roli

2018

Pattern Recognition

1,064

900

View full text Add to dashboard Cite

Learning-based pattern classifiers, including deep networks, have shown impressive performance in several application domains, ranging from computer vision to cybersecurity. However, it has also been shown that adversarial input perturbations carefully crafted either at training or at test time can easily subvert their predictions. The vulnerability of machine learning to such wild patterns (also referred to as adversarial examples), along with the design of suitable countermeasures, have been investigated in the research field of adversarial machine learning. In this work, we provide a thorough overview of the evolution of this research area over the last ten years and beyond, starting from pioneering, earlier work on the security of non-deep learning algorithms up to more recent work aimed to understand the security properties of deep learning algorithms, in the context of computer vision and cybersecurity tasks. We report interesting connections between these apparently-different lines of work, highlighting common misconceptions related to the security evaluation of machine-learning algorithms. We review the main threat models and attacks defined to this end, and discuss the main limitations of current work, along with the corresponding future challenges towards the design of more secure learning algorithms. (Fabio Roli) 1 The term malware (short for malicious software) is normally used to refer to harmful computer programs in general, including computer viruses, worms, ransomware, spyware, etc.2 More than 150 papers on this subject were published on ArXiv only in the last two years.

show abstract

Section: Security-by-design Defenses Against White-box Attacksmentioning

confidence: 99%

Wild patterns: Ten years after the rise of adversarial machine learning

Biggio

Roli

2018

Pattern Recognition

1,064

900

View full text Add to dashboard Cite

show abstract

“…Ran et al used this technology for robust face recognition [49] and robust feature extraction [50]. Xu et al proposed a robust rescaled hinge loss SVM [51]. Here, we use the HQ optimization method to improve CSLINEX-SVM and simplify its dual problem for expanding the incremental algorithm.…”

Section: A Hq Optimization For Cslinex-svmmentioning

confidence: 99%

Incremental Cost-Sensitive Support Vector Machine With Linear-Exponential Loss

Zhao

Wang

et al. 2020

IEEE Access

View full text Add to dashboard Cite

Incremental learning or online learning as a branch of machine learning has attracted more attention recently. For large-scale problems and dynamic data problem, incremental learning overwhelms batch learning, because of its efficient treatment for new data. However, class imbalance problem, which always appears in online classification brings a considerable challenge for incremental learning. The serious class imbalance problem may directly lead to a useless learning system. Cost-sensitive learning is an important learning paradigm for class imbalance problems and widely used in many applications. In this paper, we propose an incremental cost-sensitive learning method to tackle the class imbalance problems in the online situation. This proposed algorithm is based on a novel cost-sensitive support vector machine, which uses the Linear-exponential (LINEX) loss to implement high cost for minority class and low cost for majority class. Using the half-quadratic optimization, we first put forward the algorithm for the costsensitive support vector machine, called CSLINEX-SVM*. Then we propose the incremental cost-sensitive algorithm, ICSL-SVM. The results of numeric experiments demonstrate that the proposed incremental algorithm outperforms some conventional batch algorithms except the proposed CSLINEX-SVM*.

show abstract

“…where σ > 0 is window width, η = (2σ 2 ) −1 > 0 is rescaled parameter, β = 1− exp(−η) −1 > 0 is normalizing constant. Inspired by the relationship between the C-loss and the LS-loss [37], the Rhinge loss is constructed as follows…”

Section: A Loss Functions In Svmmentioning

confidence: 99%

“…The insufficiency of the study on imbalanced noisy classification of TWSVMs makes it worthwhile to adopt bounded, nonconvex loss function to pursue better performance [34]- [36]. In this paper, we introduce the rescaled hinge (Rhinge) loss [37] and give its properties, then further propose a novel improved TWSVM with monotonic, bounded, nonconvex Rhinge loss, termed RTBSVM. Owing to the adjustable parameters for each proximal hyperplane and the robustness of Rhinge loss, RTBSVM is better at tackling the imbalanced noisy classification problem.…”

Section: Introductionmentioning

confidence: 99%

Robust Rescaled Hinge Loss Twin Support Vector Machine for Imbalanced Noisy Classification

et al. 2019

View full text Add to dashboard Cite

Support vector machine (SVM) and twin SVM (TWSVM) are sensitive to the noisy classification, due to the unlimited measures in their losses, especially for imbalanced classification problem. In this paper, by combining the advantages of the correntropy induced loss function (C-Loss) and the hinge loss function (hinge loss), we introduce the rescaled hinge loss function (Rhinge loss), which is a monotonic, bounded, and nonconvex loss, into TWSVM for imbalanced noisy classification, called RTBSVM. We show that the Rhinge loss could approximate the hard margin loss and the hinge loss by adjusting the rescaled parameter, and further, our RTBSVM could improve the stability and performance of TWSVM and it is effective for imbalanced noisy classification. The experimental results show that our method performs better than the compared TWSVMs and robust SVMs on the imbalanced noisy classification.

show abstract

Robust support vector machines based on the rescaled hinge loss function

Cited by 94 publications

References 20 publications

Wild patterns: Ten years after the rise of adversarial machine learning

Wild patterns: Ten years after the rise of adversarial machine learning

Incremental Cost-Sensitive Support Vector Machine With Linear-Exponential Loss

Robust Rescaled Hinge Loss Twin Support Vector Machine for Imbalanced Noisy Classification

Contact Info

Product

Resources

About