Interpretable machine learning with reject option

Brinkrolf, Johannes; Hammer, Barbara

doi:10.1515/auto-2017-0123

Cited by 15 publications

(5 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…( 36) does, but is easier to optimize overhowever, note that by using a surrogate instead of the original constraint Eq. (36) we give up closeness which, in our opinion, would be acceptable if the solutions stay somewhat close to each other 6 . We try out and compare both approaches in the experiments (see Section 4).…”

Section: Distance To Decision Boundarymentioning

confidence: 67%

See 1 more Smart Citation

Explaining Reject Options of Learning Vector Quantization Classifiers

Artelt¹,

Brinkrolf²,

Visser³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

While machine learning models are usually assumed to always output a prediction, there also exist extensions in the form of reject options which allow the model to reject inputs where only a prediction with an unacceptably low certainty would be possible. With the ongoing rise of eXplainable AI, a lot of methods for explaining model predictions have been developed. However, understanding why a given input was rejected, instead of being classified by the model, is also of interest. Surprisingly, explanations of rejects have not been considered so far. We propose to use counterfactual explanations for explaining rejects and investigate how to efficiently compute counterfactual explanations of different reject options for an important class of models, namely prototypebased classifiers such as learning vector quantization models.

show abstract

Section: Distance To Decision Boundarymentioning

confidence: 67%

“…At the same time, RelSim gets close to 0 if the sample is far away -i.e. both distances are large [6].…”

Section: Reject Optionsmentioning

confidence: 93%

Explaining Reject Options of Learning Vector Quantization Classifiers

Artelt¹,

Brinkrolf²,

Visser³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…the classifier outputs a specific warning in the case of outliers or settings close to a decision boundary. It has recently been shown that such technology can mediate the problem of adversarial examples for some intuitive classifiers [28]. It remains a matter of research in how far this technology can also improve security for deep neural networks.…”

Section: Discussionmentioning

confidence: 99%

Generation of Adversarial Examples to Prevent Misclassification of Deep Neural Network based Condition Monitoring Systems for Cyber-Physical Production Systems

Specht

Otto

Niggemann

et al. 2018

2018 IEEE 16th International Conference on Industrial Informatics (INDIN)

Self Cite

View full text Add to dashboard Cite

Deep neural network based condition monitoring systems are used to detect system failures of cyber-physical production systems. However, a vulnerability of deep neural networks are adversarial examples. They are manipulated inputs, e.g. process data, with the ability to mislead a deep neural network into misclassification. Adversarial example attacks can manipulate the physical production process of a cyber-physical production system without being recognized by the condition monitoring system. Manipulation of the physical process poses a serious threat for production systems and employees. This paper introduces CyberProtect, a novel approach to prevent misclassification caused by adversarial example attacks. CyberProtect generates adversarial examples and uses them to retrain deep neural networks. This results in a hardened deep neural network with a significant reduced misclassification rate. The proposed countermeasure increases the classification rate from 20% to 82%, as proved by empirical results.

show abstract

“…KMeans-Proxy. Through observations on our generalizability results, we present a simple 'reject option' [4,15,16] for fake news detectors, called KMeans-Proxy. KMeans-Proxy is based on KMeans clustering, and is inspired by research into proxy losses [30,47] and foundation models [3,7,33].…”

Section: Introductionmentioning

confidence: 99%

Evaluating Generalizability of Fine-Tuned Models for Fake News Detection

Suprem¹,

Pu²

2022

Preprint

View full text Add to dashboard Cite

The Covid-19 pandemic has caused a dramatic and parallel rise in dangerous misinformation, denoted an 'infodemic' by the CDC and WHO. Misinformation tied to the Covid-19 infodemic changes continuously; this can lead to performance degradation of fine-tuned models due to concept drift. Degredation can be mitigated if models generalize well-enough to capture some cyclical aspects of drifted data. In this paper, we explore generalizability of pre-trained and fine-tuned fake news detectors across 9 fake news datasets. We show that existing models often overfit on their training dataset and have poor performance on unseen data. However, on some subsets of unseen data that overlap with training data, models have higher accuracy. Based on this observation, we also present KMeans-Proxy, a fast and effective method based on K-Means clustering for quickly identifying these overlapping subsets of unseen data. KMeans-Proxy improves generalizability on unseen fake news datasets by 0.1-0.2 f1-points across datasets. We present both our generalizability experiments as well as KMeans-Proxy to further research in tackling the fake news problem.

show abstract

Interpretable machine learning with reject option

Cited by 15 publications

References 21 publications

Explaining Reject Options of Learning Vector Quantization Classifiers

Explaining Reject Options of Learning Vector Quantization Classifiers

Generation of Adversarial Examples to Prevent Misclassification of Deep Neural Network based Condition Monitoring Systems for Cyber-Physical Production Systems

Evaluating Generalizability of Fine-Tuned Models for Fake News Detection

Contact Info

Product

Resources

About