FSL-BM: Fuzzy Supervised Learning with Binary Meta-Feature for Classification

Kowsari, Kamran; Bari, Nima; Vichr, Roman; Goodarzi, Farhad A.

doi:10.1007/978-3-030-03405-4_46

Cited by 5 publications

(4 citation statements)

References 39 publications

(42 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this sense it can even be viewed as a kind of mining [10] of preferences from raw decision data. Please also note that the algorithm is neither supervised [14] nor initial explicit knowledge about the preferences is given [11].…”

Section: Ifmentioning

confidence: 99%

Learning of Consistent Preference Relations for Decision Making and Optimization in Context of Interacting Goals

Felix¹

2021

Atlantis Studies in Uncertainty Modelling

View full text Add to dashboard Cite

In former papers a decision and optimization algorithm based on interactions between goals was introduced and it was shown how it solves relevant real-world problems. However, the algorithm assumed preference information on the goals as initially given. In this paper we describe a learning algorithm that derives such preference information from decision and optimization input data. It is shown how the former algorithm is improved and how this improvement links the algorithm to reinforcement-based machine learning optimization.

show abstract

Section: Ifmentioning

confidence: 99%

Learning of Consistent Preference Relations for Decision Making and Optimization in Context of Interacting Goals

Felix¹

2021

Atlantis Studies in Uncertainty Modelling

View full text Add to dashboard Cite

show abstract

“…However, these methods are not semi-supervised and therefore depend on large labeled datasets. A variety of methods was proposed to handle fuzzy data in a semi-supervised learning approach [33,27,14]. These methods use lower-dimensional features spaces in contrast to images as input.…”

Section: Related Workmentioning

confidence: 99%

Beyond Cats and Dogs: Semi-supervised Classification of fuzzy labels with overclustering

Schmarje,

Brünger,

Santarossa

et al. 2020

Preprint

View full text Add to dashboard Cite

A long-standing issue with deep learning is the need for large and consistently labeled datasets. Although the current research in semi-supervised learning can decrease the required amount of annotated data by a factor of 10 or even more, this line of research still uses distinct classes like cats and dogs. However, in the real-world we often encounter problems where different experts have different opinions, thus producing fuzzy labels. We propose a novel framework for handling semi-supervised classifications of such fuzzy labels. Our framework is based on the idea of overclustering to detect substructures in these fuzzy labels. We propose a novel loss to improve the overclustering capability of our framework and show on the common image classification dataset STL-10 that it is faster and has better overclustering performance than previous work. On a real-world plankton dataset, we illustrate the benefit of overclustering for fuzzy labels and show that we beat previous state-of-the-art semisupervised methods. Moreover, we acquire 5 to 10% more consistent predictions of substructures. 1 * Corresponding Author 1 The source code, the datasets and supplementary material will be released soon. Please contact the corresponding author for any questions.

show abstract

“…However, these methods are not semi-supervised and therefore depend on large labeled datasets. A variety of methods was proposed to handle data in a semi-supervised learning approach [ 33 , 34 , 35 ]. These methods use lower-dimensional features spaces in contrast to images as input.…”

Section: Introductionmentioning

confidence: 99%

Fuzzy Overclustering: Semi-Supervised Classification of Fuzzy Labels with Overclustering and Inverse Cross-Entropy

Schmarje

Brünger

Santarossa

et al. 2021

Sensors

View full text Add to dashboard Cite

Deep learning has been successfully applied to many classification problems including underwater challenges. However, a long-standing issue with deep learning is the need for large and consistently labeled datasets. Although current approaches in semi-supervised learning can decrease the required amount of annotated data by a factor of 10 or even more, this line of research still uses distinct classes. For underwater classification, and uncurated real-world datasets in general, clean class boundaries can often not be given due to a limited information content in the images and transitional stages of the depicted objects. This leads to different experts having different opinions and thus producing fuzzy labels which could also be considered ambiguous or divergent. We propose a novel framework for handling semi-supervised classifications of such fuzzy labels. It is based on the idea of overclustering to detect substructures in these fuzzy labels. We propose a novel loss to improve the overclustering capability of our framework and show the benefit of overclustering for fuzzy labels. We show that our framework is superior to previous state-of-the-art semi-supervised methods when applied to real-world plankton data with fuzzy labels. Moreover, we acquire 5 to 10% more consistent predictions of substructures.

show abstract

FSL-BM: Fuzzy Supervised Learning with Binary Meta-Feature for Classification

Cited by 5 publications

References 39 publications

Learning of Consistent Preference Relations for Decision Making and Optimization in Context of Interacting Goals

Learning of Consistent Preference Relations for Decision Making and Optimization in Context of Interacting Goals

Beyond Cats and Dogs: Semi-supervised Classification of fuzzy labels with overclustering

Fuzzy Overclustering: Semi-Supervised Classification of Fuzzy Labels with Overclustering and Inverse Cross-Entropy

Contact Info

Product

Resources

About