2014
DOI: 10.1016/j.ins.2013.07.030
|View full text |Cite
|
Sign up to set email alerts
|

Reliable classification: Learning classifiers that distinguish aleatoric and epistemic uncertainty

Abstract: A proper representation of the uncertainty involved in a prediction is an important prerequisite for the acceptance of machine learning and decision support technology in safety-critical application domains such as medical diagnosis. Despite the existence of various probabilistic approaches in these fields, there is arguably no method that is able to distinguish between two very different sources of uncertainty: aleatoric uncertainty, which is due to statistical variability and effects that are inherently rand… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

3
99
0

Year Published

2015
2015
2024
2024

Publication Types

Select...
4
4
1

Relationship

2
7

Authors

Journals

citations
Cited by 114 publications
(102 citation statements)
references
References 16 publications
3
99
0
Order By: Relevance
“…More recently, this distinction has also received attention in machine learning, where the "agent" is a learning algorithm [18]. In particular, a distinction between aleatoric and epistemic uncertainty has been advocated in the literature on deep learning [6], where the limited awareness of neural networks of their own competence has been demonstrated quite nicely.…”
Section: Introductionmentioning
confidence: 99%
“…More recently, this distinction has also received attention in machine learning, where the "agent" is a learning algorithm [18]. In particular, a distinction between aleatoric and epistemic uncertainty has been advocated in the literature on deep learning [6], where the limited awareness of neural networks of their own competence has been demonstrated quite nicely.…”
Section: Introductionmentioning
confidence: 99%
“…According to [95], one can distinguish aleatoric and epistemic uncertainty in the classifier. In the context of LVQ, aleatoric uncertainty is due to randomness in data generating scheme whereas epimistic uncertainty is related to the lack of knowledge.…”
Section: Reject or Classify -Secure Classificationmentioning
confidence: 99%
“…The term relates to epistemic uncertainty (as opposed to aleatoric uncertainty) [10] and is due to our mistreatment of the evidence X with respect to the ideal model.…”
Section: Ideal Scores Q and The Decomposition L = El + Ilmentioning
confidence: 99%
“…This type of uncertainty is called aleatoric [10] so the loss could also be called aleatoric loss. It is the loss of the optimal model and equals zero only if the attributes of the instance X provide enough information to uniquely determine the right label Y (with probability 1).…”
Section: Ideal Scores Q and The Decomposition L = El + Ilmentioning
confidence: 99%