Adaptive tangent distances in generalized learning vector quantization for transformation and distortion invariant classification learning

Saralajew, Sascha; Villmann, Thomas

doi:10.1109/ijcnn.2016.7727534

Cited by 14 publications

(12 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This was also shown in [10], where the accuracy of the Madry model is significantly lower when considering a threshold t ∞ > 0.3. 9 Furthermore, the Madry model has outstanding robustness scores for gradient-based attacks in general. We accredit this effect to potential obfuscation of gradients as a side-effect of the adversarial training procedure.…”

Section: Hypothesis Margin Maximization In the Input Space Produces Rmentioning

confidence: 99%

“…For the Generalized LVQ (GLVQ) [6], considered as a differentiable cost function based variant of LVQ, robustness is theoretically anticipated because it maximizes the hypothesis margin in the input space [7]. This changes if the squared Euclidean distance in GLVQ is replaced by adaptive dissimilarity measures such as in Generalized Matrix LVQ (GMLVQ) [8] or Generalized Tangent LVQ (GTLVQ) [9]. They first apply a projection and measure the dissimilarity in the corresponding projection space, also denoted as feature space.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Robustness of Generalized Learning Vector Quantization Models Against Adversarial Attacks

Saralajew¹,

Holdijk

Rees³

et al. 2019

Advances in Intelligent Systems and Computing

Self Cite

View full text Add to dashboard Cite

Adversarial attacks and the development of (deep) neural networks robust against them are currently two widely researched topics. The robustness of Learning Vector Quantization (LVQ) models against adversarial attacks has however not yet been studied to the same extent. We therefore present an extensive evaluation of three LVQ models: Generalized LVQ, Generalized Matrix LVQ and Generalized Tangent LVQ. The evaluation suggests that both Generalized LVQ and Generalized Tangent LVQ have a high base robustness, on par with the current state-of-the-art in robust neural network methods. In contrast to this, Generalized Matrix LVQ shows a high susceptibility to adversarial attacks, scoring consistently behind all other models. Additionally, our numerical evaluation indicates that increasing the number of prototypes per class improves the robustness of the models.

show abstract

Section: Hypothesis Margin Maximization In the Input Space Produces Rmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Robustness of Generalized Learning Vector Quantization Models Against Adversarial Attacks

Saralajew¹,

Holdijk

Rees³

et al. 2019

Advances in Intelligent Systems and Computing

Self Cite

View full text Add to dashboard Cite

show abstract

“…Recent developments include tangent metrics to deal with invariances of data regarding classification, e.g. rotations of objects in images [51,52].…”

Section: Beyond the Euclidean World -Glvq With Non-standard Dissimilamentioning

confidence: 99%

“…Recently, GLVQ was extended also to deal with this problem. More specifically, GLVQ is provided wih an adaptive tangent metric (GTLVQ) allowing to determine the respective invariances/transformations automatically during classification learning following the same principle as relevance learning [51,52].…”

Section: Relevance Learning and Related Variantsmentioning

confidence: 99%

Can Learning Vector Quantization be an Alternative to SVM and Deep Learning? - Recent Trends and Advanced Variants of Learning Vector Quantization for Classification Learning

Villmann

Bohnsack

Kaden

2016

Journal of Artificial Intelligence and Soft Computing Research

Self Cite

View full text Add to dashboard Cite

Learning vector quantization (LVQ) is one of the most powerful approaches for prototype based classification of vector data, intuitively introduced by Kohonen. The prototype adaptation scheme relies on its attraction and repulsion during the learning providing an easy geometric interpretability of the learning as well as of the classification decision scheme. Although deep learning architectures and support vector classifiers frequently achieve comparable or even better results, LVQ models are smart alternatives with low complexity and computational costs making them attractive for many industrial applications like intelligent sensor systems or advanced driver assistance systems.Nowadays, the mathematical theory developed for LVQ delivers sufficient justification of the algorithm making it an appealing alternative to other approaches like support vector machines and deep learning techniques.This review article reports current developments and extensions of LVQ starting from the generalized LVQ (GLVQ), which is known as the most powerful cost function based realization of the original LVQ. The cost function minimized in GLVQ is an soft-approximation of the standard classification error allowing gradient descent learning techniques. The GLVQ variants considered in this contribution, cover many aspects like bordersensitive learning, application of non-Euclidean metrics like kernel distances or divergences, relevance learning as well as optimization of advanced statistical classification quality measures beyond the accuracy including sensitivity and specificity or area under the ROC-curve.According to these topics, the paper highlights the basic motivation for these variants and extensions together with the mathematical prerequisites and treatments for integration into the standard GLVQ scheme and compares them to other machine learning approaches. For detailed description and mathematical theory behind all, the reader is referred to the respective original articles.Thus, the intention of the paper is to provide a comprehensive overview of the stateof-the-art serving as a starting point to search for an appropriate LVQ variant in case of a given specific classification problem as well as a reference to recently developed variants and improvements of the basic GLVQ scheme.

show abstract

“…The Generalized Tangent Learning Vector Quantization (GTLVQ) [7] models the classification boundary implicitly by approximating the local data manifold structure via affine subspaces (tangent space approximations). Employing the GTLVQ as domain tangent discriminator is favorable to address the above issues because (i) provides a locally invariant and reliable model in the adaptation process by subspace and online learning, (ii) can capture multi-modal manifold structures, and (iii) provides an interpretable model by visualizing points from the affine subspace to verify the adaptation process.…”

Section: Introductionmentioning

confidence: 99%

Domain Adversarial Tangent Learning Towards Interpretable Domain Adaptation

Raab¹,

Saralajew²,

Schleif³

2021

ESANN 2021 Proceedings

Self Cite

View full text Add to dashboard Cite

Deep learning struggles to generalize well to an unseen target domain of interest. Current domain adaptation methods simultaneously learn a classifier and an adversarial game for invariant representations but inadequately align local structures, while the underlying process is hard to interpret. We propose a new interpretable adversarial domain architecture, matching local manifold approximations across domains. Evaluated against related networks, the approach is competitive, while the adaptation process can be visually verified.

show abstract

Adaptive tangent distances in generalized learning vector quantization for transformation and distortion invariant classification learning

Cited by 14 publications

References 21 publications

Robustness of Generalized Learning Vector Quantization Models Against Adversarial Attacks

Robustness of Generalized Learning Vector Quantization Models Against Adversarial Attacks

Can Learning Vector Quantization be an Alternative to SVM and Deep Learning? - Recent Trends and Advanced Variants of Learning Vector Quantization for Classification Learning

Domain Adversarial Tangent Learning Towards Interpretable Domain Adaptation

Contact Info

Product

Resources

About