---
_id: '44512'
abstract:
- lang: eng
  text: "For open world applications, deep neural networks (DNNs) need to be aware
    of\r\npreviously unseen data and adaptable to evolving environments. Furthermore,
    it\r\nis desirable to detect and learn novel classes which are not included in
    the\r\nDNNs underlying set of semantic classes in an unsupervised fashion. The
    method\r\nproposed in this article builds upon anomaly detection to retrieve\r\nout-of-distribution
    (OoD) data as candidates for new classes. We thereafter\r\nextend the DNN by $k$
    empty classes and fine-tune it on the OoD data samples.\r\nTo this end, we introduce
    two loss functions, which 1) entice the DNN to assign\r\nOoD samples to the empty
    classes and 2) to minimize the inner-class feature\r\ndistances between them.
    Thus, instead of ground truth which contains labels for\r\nthe different novel
    classes, the DNN obtains a single OoD label together with a\r\ndistance matrix,
    which is computed in advance. We perform several experiments\r\nfor image classification
    and semantic segmentation, which demonstrate that a\r\nDNN can extend its own
    semantic space by multiple classes without having access\r\nto ground truth."
author:
- first_name: Svenja
  full_name: Uhlemeyer, Svenja
  last_name: Uhlemeyer
- first_name: Julian
  full_name: Lienen, Julian
  id: '44040'
  last_name: Lienen
- first_name: Eyke
  full_name: Hüllermeier, Eyke
  id: '48129'
  last_name: Hüllermeier
- first_name: Hanno
  full_name: Gottschalk, Hanno
  last_name: Gottschalk
citation:
  ama: Uhlemeyer S, Lienen J, Hüllermeier E, Gottschalk H. Detecting Novelties with
    Empty Classes. <i>arXiv:230500983</i>. Published online 2023.
  apa: Uhlemeyer, S., Lienen, J., Hüllermeier, E., &#38; Gottschalk, H. (2023). Detecting
    Novelties with Empty Classes. In <i>arXiv:2305.00983</i>.
  bibtex: '@article{Uhlemeyer_Lienen_Hüllermeier_Gottschalk_2023, title={Detecting
    Novelties with Empty Classes}, journal={arXiv:2305.00983}, author={Uhlemeyer,
    Svenja and Lienen, Julian and Hüllermeier, Eyke and Gottschalk, Hanno}, year={2023}
    }'
  chicago: Uhlemeyer, Svenja, Julian Lienen, Eyke Hüllermeier, and Hanno Gottschalk.
    “Detecting Novelties with Empty Classes.” <i>ArXiv:2305.00983</i>, 2023.
  ieee: S. Uhlemeyer, J. Lienen, E. Hüllermeier, and H. Gottschalk, “Detecting Novelties
    with Empty Classes,” <i>arXiv:2305.00983</i>. 2023.
  mla: Uhlemeyer, Svenja, et al. “Detecting Novelties with Empty Classes.” <i>ArXiv:2305.00983</i>,
    2023.
  short: S. Uhlemeyer, J. Lienen, E. Hüllermeier, H. Gottschalk, ArXiv:2305.00983
    (2023).
date_created: 2023-05-05T11:37:00Z
date_updated: 2023-05-05T11:39:10Z
external_id:
  arxiv:
  - '2305.00983'
language:
- iso: eng
main_file_link:
- open_access: '1'
  url: https://arxiv.org/pdf/2305.00983.pdf
oa: '1'
publication: arXiv:2305.00983
status: public
title: Detecting Novelties with Empty Classes
type: preprint
user_id: '44040'
year: '2023'
...
---
_id: '31880'
abstract:
- lang: eng
  text: The notion of neural collapse refers to several emergent phenomena that have
    been empirically observed across various canonical classification problems. During
    the terminal phase of training a deep neural network, the feature embedding of
    all examples of the same class tend to collapse to a single representation, and
    the features of different classes tend to separate as much as possible. Neural
    collapse is often studied through a simplified model, called the unconstrained
    feature representation, in which the model is assumed to have "infinite expressivity"
    and can map each data point to any arbitrary representation. In this work, we
    propose a more realistic variant of the unconstrained feature representation that
    takes the limited expressivity of the network into account. Empirical evidence
    suggests that the memorization of noisy data points leads to a degradation (dilation)
    of the neural collapse. Using a model of the memorization-dilation (M-D) phenomenon,
    we show one mechanism by which different losses lead to different performances
    of the trained network on noisy data. Our proofs reveal why label smoothing, a
    modification of cross-entropy empirically observed to produce a regularization
    effect, leads to improved generalization in classification tasks.
author:
- first_name: Duc Anh
  full_name: Nguyen, Duc Anh
  last_name: Nguyen
- first_name: Ron
  full_name: Levie, Ron
  last_name: Levie
- first_name: Julian
  full_name: Lienen, Julian
  id: '44040'
  last_name: Lienen
- first_name: Gitta
  full_name: Kutyniok, Gitta
  last_name: Kutyniok
- first_name: Eyke
  full_name: Hüllermeier, Eyke
  id: '48129'
  last_name: Hüllermeier
citation:
  ama: 'Nguyen DA, Levie R, Lienen J, Kutyniok G, Hüllermeier E. Memorization-Dilation:
    Modeling Neural Collapse Under Noise. In: <i>International Conference on Learning
    Representations, ICLR</i>. ; 2023.'
  apa: 'Nguyen, D. A., Levie, R., Lienen, J., Kutyniok, G., &#38; Hüllermeier, E.
    (2023). Memorization-Dilation: Modeling Neural Collapse Under Noise. <i>International
    Conference on Learning Representations, ICLR</i>. International Conference on
    Learning Representations, ICLR, Kigali, Ruanda.'
  bibtex: '@inproceedings{Nguyen_Levie_Lienen_Kutyniok_Hüllermeier_2023, title={Memorization-Dilation:
    Modeling Neural Collapse Under Noise}, booktitle={International Conference on
    Learning Representations, ICLR}, author={Nguyen, Duc Anh and Levie, Ron and Lienen,
    Julian and Kutyniok, Gitta and Hüllermeier, Eyke}, year={2023} }'
  chicago: 'Nguyen, Duc Anh, Ron Levie, Julian Lienen, Gitta Kutyniok, and Eyke Hüllermeier.
    “Memorization-Dilation: Modeling Neural Collapse Under Noise.” In <i>International
    Conference on Learning Representations, ICLR</i>, 2023.'
  ieee: 'D. A. Nguyen, R. Levie, J. Lienen, G. Kutyniok, and E. Hüllermeier, “Memorization-Dilation:
    Modeling Neural Collapse Under Noise,” presented at the International Conference
    on Learning Representations, ICLR, Kigali, Ruanda, 2023.'
  mla: 'Nguyen, Duc Anh, et al. “Memorization-Dilation: Modeling Neural Collapse Under
    Noise.” <i>International Conference on Learning Representations, ICLR</i>, 2023.'
  short: 'D.A. Nguyen, R. Levie, J. Lienen, G. Kutyniok, E. Hüllermeier, in: International
    Conference on Learning Representations, ICLR, 2023.'
conference:
  location: Kigali, Ruanda
  name: International Conference on Learning Representations, ICLR
date_created: 2022-06-14T14:48:36Z
date_updated: 2023-06-29T09:14:26Z
language:
- iso: eng
main_file_link:
- open_access: '1'
  url: https://arxiv.org/abs/2206.05530
oa: '1'
publication: International Conference on Learning Representations, ICLR
status: public
title: 'Memorization-Dilation: Modeling Neural Collapse Under Noise'
type: conference
user_id: '44040'
year: '2023'
...
---
_id: '45911'
abstract:
- lang: eng
  text: "Label noise poses an important challenge in machine learning, especially
    in\r\ndeep learning, in which large models with high expressive power dominate
    the\r\nfield. Models of that kind are prone to memorizing incorrect labels, thereby\r\nharming
    generalization performance. Many methods have been proposed to address\r\nthis
    problem, including robust loss functions and more complex label correction\r\napproaches.
    Robust loss functions are appealing due to their simplicity, but\r\ntypically
    lack flexibility, while label correction usually adds substantial\r\ncomplexity
    to the training setup. In this paper, we suggest to address the\r\nshortcomings
    of both methodologies by \"ambiguating\" the target information,\r\nadding additional,
    complementary candidate labels in case the learner is not\r\nsufficiently convinced
    of the observed training label. More precisely, we\r\nleverage the framework of
    so-called superset learning to construct set-valued\r\ntargets based on a confidence
    threshold, which deliver imprecise yet more\r\nreliable beliefs about the ground-truth,
    effectively helping the learner to\r\nsuppress the memorization effect. In an
    extensive empirical evaluation, our\r\nmethod demonstrates favorable learning
    behavior on synthetic and real-world\r\nnoise, confirming the effectiveness in
    detecting and correcting erroneous\r\ntraining labels."
author:
- first_name: Julian
  full_name: Lienen, Julian
  id: '44040'
  last_name: Lienen
- first_name: Eyke
  full_name: Hüllermeier, Eyke
  id: '48129'
  last_name: Hüllermeier
citation:
  ama: Lienen J, Hüllermeier E. Mitigating Label Noise through Data Ambiguation. <i>arXiv:230513764</i>.
    Published online 2023.
  apa: Lienen, J., &#38; Hüllermeier, E. (2023). Mitigating Label Noise through Data
    Ambiguation. In <i>arXiv:2305.13764</i>.
  bibtex: '@article{Lienen_Hüllermeier_2023, title={Mitigating Label Noise through
    Data Ambiguation}, journal={arXiv:2305.13764}, author={Lienen, Julian and Hüllermeier,
    Eyke}, year={2023} }'
  chicago: Lienen, Julian, and Eyke Hüllermeier. “Mitigating Label Noise through Data
    Ambiguation.” <i>ArXiv:2305.13764</i>, 2023.
  ieee: J. Lienen and E. Hüllermeier, “Mitigating Label Noise through Data Ambiguation,”
    <i>arXiv:2305.13764</i>. 2023.
  mla: Lienen, Julian, and Eyke Hüllermeier. “Mitigating Label Noise through Data
    Ambiguation.” <i>ArXiv:2305.13764</i>, 2023.
  short: J. Lienen, E. Hüllermeier, ArXiv:2305.13764 (2023).
date_created: 2023-07-09T11:25:48Z
date_updated: 2023-07-09T11:26:21Z
external_id:
  arxiv:
  - '2305.13764'
language:
- iso: eng
main_file_link:
- open_access: '1'
  url: https://arxiv.org/abs/2305.13764
oa: '1'
publication: arXiv:2305.13764
status: public
title: Mitigating Label Noise through Data Ambiguation
type: preprint
user_id: '44040'
year: '2023'
...
---
_id: '34542'
author:
- first_name: Andrea
  full_name: Campagner, Andrea
  last_name: Campagner
- first_name: Julian
  full_name: Lienen, Julian
  id: '44040'
  last_name: Lienen
- first_name: Eyke
  full_name: Hüllermeier, Eyke
  id: '48129'
  last_name: Hüllermeier
- first_name: Davide
  full_name: Ciucci, Davide
  last_name: Ciucci
citation:
  ama: 'Campagner A, Lienen J, Hüllermeier E, Ciucci D. Scikit-Weak: A Python Library
    for Weakly Supervised Machine Learning. In: <i>Lecture Notes in Computer Science</i>.
    Vol 13633. Springer; 2022:57-70.'
  apa: 'Campagner, A., Lienen, J., Hüllermeier, E., &#38; Ciucci, D. (2022). Scikit-Weak:
    A Python Library for Weakly Supervised Machine Learning. <i>Lecture Notes in Computer
    Science</i>, <i>13633</i>, 57–70.'
  bibtex: '@inproceedings{Campagner_Lienen_Hüllermeier_Ciucci_2022, title={Scikit-Weak:
    A Python Library for Weakly Supervised Machine Learning}, volume={13633}, booktitle={Lecture
    Notes in Computer Science}, publisher={Springer}, author={Campagner, Andrea and
    Lienen, Julian and Hüllermeier, Eyke and Ciucci, Davide}, year={2022}, pages={57–70}
    }'
  chicago: 'Campagner, Andrea, Julian Lienen, Eyke Hüllermeier, and Davide Ciucci.
    “Scikit-Weak: A Python Library for Weakly Supervised Machine Learning.” In <i>Lecture
    Notes in Computer Science</i>, 13633:57–70. Springer, 2022.'
  ieee: 'A. Campagner, J. Lienen, E. Hüllermeier, and D. Ciucci, “Scikit-Weak: A Python
    Library for Weakly Supervised Machine Learning,” in <i>Lecture Notes in Computer
    Science</i>, Suzhou, China, 2022, vol. 13633, pp. 57–70.'
  mla: 'Campagner, Andrea, et al. “Scikit-Weak: A Python Library for Weakly Supervised
    Machine Learning.” <i>Lecture Notes in Computer Science</i>, vol. 13633, Springer,
    2022, pp. 57–70.'
  short: 'A. Campagner, J. Lienen, E. Hüllermeier, D. Ciucci, in: Lecture Notes in
    Computer Science, Springer, 2022, pp. 57–70.'
conference:
  end_date: 2022-11-14
  location: Suzhou, China
  name: International Joint Conference on Rough Sets
  start_date: 2022-11-11
date_created: 2022-12-19T09:34:35Z
date_updated: 2022-12-19T09:34:44Z
intvolume: '     13633'
language:
- iso: eng
page: 57-70
publication: Lecture Notes in Computer Science
publisher: Springer
status: public
title: 'Scikit-Weak: A Python Library for Weakly Supervised Machine Learning'
type: conference
user_id: '44040'
volume: 13633
year: '2022'
...
---
_id: '31545'
abstract:
- lang: eng
  text: Knowledge graph embedding research has mainly focused on learning continuous
    representations of entities and relations tailored towards the link prediction
    problem. Recent results indicate an ever increasing predictive ability of current
    approaches on benchmark datasets. However, this effectiveness often comes with
    the cost of over-parameterization and increased computationally complexity. The
    former induces extensive hyperparameter optimization to mitigate malicious overfitting.
    The latter magnifies the importance of winning the hardware lottery. Here, we
    investigate a remedy for the first problem. We propose a technique based on Kronecker
    decomposition to reduce the number of parameters in a knowledge graph embedding
    model, while retaining its expressiveness. Through Kronecker decomposition, large
    embedding matrices are split into smaller embedding matrices during the training
    process. Hence, embeddings of knowledge graphs are not plainly retrieved but reconstructed
    on the fly. The decomposition ensures that elementwise interactions between three
    embedding vectors are extended with interactions within each embedding vector.
    This implicitly reduces redundancy in embedding vectors and encourages feature
    reuse. To quantify the impact of applying Kronecker decomposition on embedding
    matrices, we conduct a series of experiments on benchmark datasets. Our experiments
    suggest that applying Kronecker decomposition on embedding matrices leads to an
    improved parameter efficiency on all benchmark datasets. Moreover, empirical evidence
    suggests that reconstructed embeddings entail robustness against noise in the
    input knowledge graph. To foster reproducible research, we provide an open-source
    implementation of our approach, including training and evaluation scripts as well
    as pre-trained models in our knowledge graph embedding framework.
author:
- first_name: Caglar
  full_name: Demir, Caglar
  id: '43817'
  last_name: Demir
- first_name: Julian
  full_name: Lienen, Julian
  id: '44040'
  last_name: Lienen
- first_name: Axel-Cyrille
  full_name: Ngonga Ngomo, Axel-Cyrille
  id: '65716'
  last_name: Ngonga Ngomo
citation:
  ama: Demir C, Lienen J, Ngonga Ngomo A-C. Kronecker Decomposition for Knowledge
    Graph Embeddings. <i>arXiv:220506560</i>. Published online 2022.
  apa: Demir, C., Lienen, J., &#38; Ngonga Ngomo, A.-C. (2022). Kronecker Decomposition
    for Knowledge Graph Embeddings. In <i>arXiv:2205.06560</i>.
  bibtex: '@article{Demir_Lienen_Ngonga Ngomo_2022, title={Kronecker Decomposition
    for Knowledge Graph Embeddings}, journal={arXiv:2205.06560}, author={Demir, Caglar
    and Lienen, Julian and Ngonga Ngomo, Axel-Cyrille}, year={2022} }'
  chicago: Demir, Caglar, Julian Lienen, and Axel-Cyrille Ngonga Ngomo. “Kronecker
    Decomposition for Knowledge Graph Embeddings.” <i>ArXiv:2205.06560</i>, 2022.
  ieee: C. Demir, J. Lienen, and A.-C. Ngonga Ngomo, “Kronecker Decomposition for
    Knowledge Graph Embeddings,” <i>arXiv:2205.06560</i>. 2022.
  mla: Demir, Caglar, et al. “Kronecker Decomposition for Knowledge Graph Embeddings.”
    <i>ArXiv:2205.06560</i>, 2022.
  short: C. Demir, J. Lienen, A.-C. Ngonga Ngomo, ArXiv:2205.06560 (2022).
date_created: 2022-05-31T07:04:36Z
date_updated: 2022-05-31T07:05:50Z
language:
- iso: eng
main_file_link:
- open_access: '1'
  url: https://arxiv.org/abs/2205.06560
oa: '1'
publication: arXiv:2205.06560
status: public
title: Kronecker Decomposition for Knowledge Graph Embeddings
type: preprint
user_id: '44040'
year: '2022'
...
---
_id: '31546'
abstract:
- lang: eng
  text: In semi-supervised learning, the paradigm of self-training refers to the idea
    of learning from pseudo-labels suggested by the learner itself. Across various
    domains, corresponding methods have proven effective and achieve state-of-the-art
    performance. However, pseudo-labels typically stem from ad-hoc heuristics, relying
    on the quality of the predictions though without guaranteeing their validity.
    One such method, so-called credal self-supervised learning, maintains pseudo-supervision
    in the form of sets of (instead of single) probability distributions over labels,
    thereby allowing for a flexible yet uncertainty-aware labeling. Again, however,
    there is no justification beyond empirical effectiveness. To address this deficiency,
    we make use of conformal prediction, an approach that comes with guarantees on
    the validity of set-valued predictions. As a result, the construction of credal
    sets of labels is supported by a rigorous theoretical foundation, leading to better
    calibrated and less error-prone supervision for unlabeled data. Along with this,
    we present effective algorithms for learning from credal self-supervision. An
    empirical study demonstrates excellent calibration properties of the pseudo-supervision,
    as well as the competitiveness of our method on several benchmark datasets.
author:
- first_name: Julian
  full_name: Lienen, Julian
  id: '44040'
  last_name: Lienen
- first_name: Caglar
  full_name: Demir, Caglar
  id: '43817'
  last_name: Demir
- first_name: Eyke
  full_name: Hüllermeier, Eyke
  id: '48129'
  last_name: Hüllermeier
citation:
  ama: Lienen J, Demir C, Hüllermeier E. Conformal Credal Self-Supervised Learning.
    <i>arXiv:220515239</i>. Published online 2022.
  apa: Lienen, J., Demir, C., &#38; Hüllermeier, E. (2022). Conformal Credal Self-Supervised
    Learning. In <i>arXiv:2205.15239</i>.
  bibtex: '@article{Lienen_Demir_Hüllermeier_2022, title={Conformal Credal Self-Supervised
    Learning}, journal={arXiv:2205.15239}, author={Lienen, Julian and Demir, Caglar
    and Hüllermeier, Eyke}, year={2022} }'
  chicago: Lienen, Julian, Caglar Demir, and Eyke Hüllermeier. “Conformal Credal Self-Supervised
    Learning.” <i>ArXiv:2205.15239</i>, 2022.
  ieee: J. Lienen, C. Demir, and E. Hüllermeier, “Conformal Credal Self-Supervised
    Learning,” <i>arXiv:2205.15239</i>. 2022.
  mla: Lienen, Julian, et al. “Conformal Credal Self-Supervised Learning.” <i>ArXiv:2205.15239</i>,
    2022.
  short: J. Lienen, C. Demir, E. Hüllermeier, ArXiv:2205.15239 (2022).
date_created: 2022-05-31T07:05:36Z
date_updated: 2022-05-31T07:05:54Z
language:
- iso: eng
main_file_link:
- open_access: '1'
  url: https://arxiv.org/abs/2205.15239
oa: '1'
publication: arXiv:2205.15239
status: public
title: Conformal Credal Self-Supervised Learning
type: preprint
user_id: '44040'
year: '2022'
...
---
_id: '27161'
author:
- first_name: Julian
  full_name: Lienen, Julian
  id: '44040'
  last_name: Lienen
- first_name: Eyke
  full_name: Hüllermeier, Eyke
  last_name: Hüllermeier
citation:
  ama: 'Lienen J, Hüllermeier E. Credal Self-Supervised Learning. In: <i>Advances
    in Neural Information Processing Systems 34: Annual Conference on Neural Information
    Processing Systems, NeurIPS</i>. ; 2021.'
  apa: 'Lienen, J., &#38; Hüllermeier, E. (2021). Credal Self-Supervised Learning.
    <i>Advances in Neural Information Processing Systems 34: Annual Conference on
    Neural Information Processing Systems, NeurIPS</i>. Annual Conference on Neural
    Information Processing Systems, NeurIPS, Online.'
  bibtex: '@inproceedings{Lienen_Hüllermeier_2021, title={Credal Self-Supervised Learning},
    booktitle={Advances in Neural Information Processing Systems 34: Annual Conference
    on Neural Information Processing Systems, NeurIPS}, author={Lienen, Julian and
    Hüllermeier, Eyke}, year={2021} }'
  chicago: 'Lienen, Julian, and Eyke Hüllermeier. “Credal Self-Supervised Learning.”
    In <i>Advances in Neural Information Processing Systems 34: Annual Conference
    on Neural Information Processing Systems, NeurIPS</i>, 2021.'
  ieee: J. Lienen and E. Hüllermeier, “Credal Self-Supervised Learning,” presented
    at the Annual Conference on Neural Information Processing Systems, NeurIPS, Online,
    2021.
  mla: 'Lienen, Julian, and Eyke Hüllermeier. “Credal Self-Supervised Learning.” <i>Advances
    in Neural Information Processing Systems 34: Annual Conference on Neural Information
    Processing Systems, NeurIPS</i>, 2021.'
  short: 'J. Lienen, E. Hüllermeier, in: Advances in Neural Information Processing
    Systems 34: Annual Conference on Neural Information Processing Systems, NeurIPS,
    2021.'
conference:
  end_date: 2021-12-14
  location: Online
  name: Annual Conference on Neural Information Processing Systems, NeurIPS
  start_date: 2021-12-06
date_created: 2021-11-05T10:27:51Z
date_updated: 2022-01-06T06:57:35Z
language:
- iso: eng
publication: 'Advances in Neural Information Processing Systems 34: Annual Conference
  on Neural Information Processing Systems, NeurIPS'
status: public
title: Credal Self-Supervised Learning
type: conference
user_id: '44040'
year: '2021'
...
---
_id: '27162'
author:
- first_name: Julian
  full_name: Lienen, Julian
  id: '44040'
  last_name: Lienen
- first_name: Nils
  full_name: Nommensen, Nils
  last_name: Nommensen
- first_name: Ralph
  full_name: Ewerth, Ralph
  last_name: Ewerth
- first_name: Eyke
  full_name: Hüllermeier, Eyke
  last_name: Hüllermeier
citation:
  ama: 'Lienen J, Nommensen N, Ewerth R, Hüllermeier E. Robust Regression for Monocular
    Depth Estimation. In: <i>13th Asian Conference on Machine Learning, ACML</i>.
    ; 2021.'
  apa: Lienen, J., Nommensen, N., Ewerth, R., &#38; Hüllermeier, E. (2021). Robust
    Regression for Monocular Depth Estimation. <i>13th Asian Conference on Machine
    Learning, ACML</i>. 13th Asian Conference on Machine Learning, ACML, Online.
  bibtex: '@inproceedings{Lienen_Nommensen_Ewerth_Hüllermeier_2021, title={Robust
    Regression for Monocular Depth Estimation}, booktitle={13th Asian Conference on
    Machine Learning, ACML}, author={Lienen, Julian and Nommensen, Nils and Ewerth,
    Ralph and Hüllermeier, Eyke}, year={2021} }'
  chicago: Lienen, Julian, Nils Nommensen, Ralph Ewerth, and Eyke Hüllermeier. “Robust
    Regression for Monocular Depth Estimation.” In <i>13th Asian Conference on Machine
    Learning, ACML</i>, 2021.
  ieee: J. Lienen, N. Nommensen, R. Ewerth, and E. Hüllermeier, “Robust Regression
    for Monocular Depth Estimation,” presented at the 13th Asian Conference on Machine
    Learning, ACML, Online, 2021.
  mla: Lienen, Julian, et al. “Robust Regression for Monocular Depth Estimation.”
    <i>13th Asian Conference on Machine Learning, ACML</i>, 2021.
  short: 'J. Lienen, N. Nommensen, R. Ewerth, E. Hüllermeier, in: 13th Asian Conference
    on Machine Learning, ACML, 2021.'
conference:
  end_date: 2021-11-19
  location: Online
  name: 13th Asian Conference on Machine Learning, ACML
  start_date: 2021-11-17
date_created: 2021-11-05T10:29:38Z
date_updated: 2022-01-06T06:57:35Z
language:
- iso: eng
publication: 13th Asian Conference on Machine Learning, ACML
status: public
title: Robust Regression for Monocular Depth Estimation
type: conference
user_id: '44040'
year: '2021'
...
---
_id: '21636'
author:
- first_name: Julian
  full_name: Lienen, Julian
  id: '44040'
  last_name: Lienen
- first_name: Eyke
  full_name: Hüllermeier, Eyke
  id: '48129'
  last_name: Hüllermeier
citation:
  ama: Lienen J, Hüllermeier E. Instance weighting through data imprecisiation. <i>International
    Journal of Approximate Reasoning</i>. 2021.
  apa: Lienen, J., &#38; Hüllermeier, E. (2021). Instance weighting through data imprecisiation.
    <i>International Journal of Approximate Reasoning</i>.
  bibtex: '@article{Lienen_Hüllermeier_2021, title={Instance weighting through data
    imprecisiation}, journal={International Journal of Approximate Reasoning}, publisher={Elsevier},
    author={Lienen, Julian and Hüllermeier, Eyke}, year={2021} }'
  chicago: Lienen, Julian, and Eyke Hüllermeier. “Instance Weighting through Data
    Imprecisiation.” <i>International Journal of Approximate Reasoning</i>, 2021.
  ieee: J. Lienen and E. Hüllermeier, “Instance weighting through data imprecisiation,”
    <i>International Journal of Approximate Reasoning</i>, 2021.
  mla: Lienen, Julian, and Eyke Hüllermeier. “Instance Weighting through Data Imprecisiation.”
    <i>International Journal of Approximate Reasoning</i>, Elsevier, 2021.
  short: J. Lienen, E. Hüllermeier, International Journal of Approximate Reasoning
    (2021).
date_created: 2021-04-20T06:48:18Z
date_updated: 2022-01-06T06:55:08Z
language:
- iso: eng
main_file_link:
- url: https://www.sciencedirect.com/science/article/pii/S0888613X21000463
publication: International Journal of Approximate Reasoning
publisher: Elsevier
status: public
title: Instance weighting through data imprecisiation
type: journal_article
user_id: '44040'
year: '2021'
...
---
_id: '21637'
author:
- first_name: Julian
  full_name: Lienen, Julian
  id: '44040'
  last_name: Lienen
- first_name: Eyke
  full_name: Hüllermeier, Eyke
  id: '48129'
  last_name: Hüllermeier
citation:
  ama: 'Lienen J, Hüllermeier E. From Label Smoothing to Label Relaxation. In: <i>Proceedings
    of the 35th AAAI Conference on Artificial Intelligence, AAAI</i>. Vol 35. AAAI
    Press; 2021:8583-8591.'
  apa: 'Lienen, J., &#38; Hüllermeier, E. (2021). From Label Smoothing to Label Relaxation.
    In <i>Proceedings of the 35th AAAI Conference on Artificial Intelligence, AAAI</i>
    (Vol. 35, pp. 8583–8591). Online: AAAI Press.'
  bibtex: '@inproceedings{Lienen_Hüllermeier_2021, title={From Label Smoothing to
    Label Relaxation}, volume={35}, number={10}, booktitle={Proceedings of the 35th
    AAAI Conference on Artificial Intelligence, AAAI}, publisher={AAAI Press}, author={Lienen,
    Julian and Hüllermeier, Eyke}, year={2021}, pages={8583–8591} }'
  chicago: Lienen, Julian, and Eyke Hüllermeier. “From Label Smoothing to Label Relaxation.”
    In <i>Proceedings of the 35th AAAI Conference on Artificial Intelligence, AAAI</i>,
    35:8583–91. AAAI Press, 2021.
  ieee: J. Lienen and E. Hüllermeier, “From Label Smoothing to Label Relaxation,”
    in <i>Proceedings of the 35th AAAI Conference on Artificial Intelligence, AAAI</i>,
    Online, 2021, vol. 35, no. 10, pp. 8583–8591.
  mla: Lienen, Julian, and Eyke Hüllermeier. “From Label Smoothing to Label Relaxation.”
    <i>Proceedings of the 35th AAAI Conference on Artificial Intelligence, AAAI</i>,
    vol. 35, no. 10, AAAI Press, 2021, pp. 8583–91.
  short: 'J. Lienen, E. Hüllermeier, in: Proceedings of the 35th AAAI Conference on
    Artificial Intelligence, AAAI, AAAI Press, 2021, pp. 8583–8591.'
conference:
  end_date: 2021-02-09
  location: Online
  name: 35th AAAI Conference on Artificial Intelligence, AAAI
  start_date: 2021-02-02
date_created: 2021-04-20T06:50:43Z
date_updated: 2022-01-06T06:55:08Z
intvolume: '        35'
issue: '10'
language:
- iso: eng
main_file_link:
- open_access: '1'
  url: https://ojs.aaai.org/index.php/AAAI/article/view/17041
oa: '1'
page: 8583-8591
publication: Proceedings of the 35th AAAI Conference on Artificial Intelligence, AAAI
publisher: AAAI Press
status: public
title: From Label Smoothing to Label Relaxation
type: conference
user_id: '44040'
volume: 35
year: '2021'
...
---
_id: '22280'
author:
- first_name: Julian
  full_name: Lienen, Julian
  id: '44040'
  last_name: Lienen
- first_name: Eyke
  full_name: Hüllermeier, Eyke
  id: '48129'
  last_name: Hüllermeier
- first_name: Ralph
  full_name: Ewerth, Ralph
  last_name: Ewerth
- first_name: Nils
  full_name: Nommensen, Nils
  last_name: Nommensen
citation:
  ama: 'Lienen J, Hüllermeier E, Ewerth R, Nommensen N. Monocular Depth Estimation
    via Listwise Ranking using the Plackett-Luce Model. In: <i>Proceedings of the
    IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR</i>. ; 2021:14595-14604.'
  apa: Lienen, J., Hüllermeier, E., Ewerth, R., &#38; Nommensen, N. (2021). Monocular
    Depth Estimation via Listwise Ranking using the Plackett-Luce Model. <i>Proceedings
    of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR</i>,
    14595–14604.
  bibtex: '@inproceedings{Lienen_Hüllermeier_Ewerth_Nommensen_2021, title={Monocular
    Depth Estimation via Listwise Ranking using the Plackett-Luce Model}, booktitle={Proceedings
    of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR},
    author={Lienen, Julian and Hüllermeier, Eyke and Ewerth, Ralph and Nommensen,
    Nils}, year={2021}, pages={14595–14604} }'
  chicago: Lienen, Julian, Eyke Hüllermeier, Ralph Ewerth, and Nils Nommensen. “Monocular
    Depth Estimation via Listwise Ranking Using the Plackett-Luce Model.” In <i>Proceedings
    of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR</i>,
    14595–604, 2021.
  ieee: J. Lienen, E. Hüllermeier, R. Ewerth, and N. Nommensen, “Monocular Depth Estimation
    via Listwise Ranking using the Plackett-Luce Model,” in <i>Proceedings of the
    IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR</i>, Online,
    2021, pp. 14595–14604.
  mla: Lienen, Julian, et al. “Monocular Depth Estimation via Listwise Ranking Using
    the Plackett-Luce Model.” <i>Proceedings of the IEEE/CVF Conference on Computer
    Vision and Pattern Recognition, CVPR</i>, 2021, pp. 14595–604.
  short: 'J. Lienen, E. Hüllermeier, R. Ewerth, N. Nommensen, in: Proceedings of the
    IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR, 2021, pp.
    14595–14604.'
conference:
  end_date: 2021-06-25
  location: Online
  name: IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR
  start_date: 2021-06-19
date_created: 2021-06-02T10:35:40Z
date_updated: 2022-01-06T06:55:29Z
language:
- iso: eng
page: 14595-14604
publication: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern
  Recognition, CVPR
status: public
title: Monocular Depth Estimation via Listwise Ranking using the Plackett-Luce Model
type: conference
user_id: '44040'
year: '2021'
...
---
_id: '22509'
abstract:
- lang: eng
  text: Self-training is an effective approach to semi-supervised learning. The key
    idea is to let the learner itself iteratively generate "pseudo-supervision" for
    unlabeled instances based on its current hypothesis. In combination with consistency
    regularization, pseudo-labeling has shown promising performance in various domains,
    for example in computer vision. To account for the hypothetical nature of the
    pseudo-labels, these are commonly provided in the form of probability distributions.
    Still, one may argue that even a probability distribution represents an excessive
    level of informedness, as it suggests that the learner precisely knows the ground-truth
    conditional probabilities. In our approach, we therefore allow the learner to
    label instances in the form of credal sets, that is, sets of (candidate) probability
    distributions. Thanks to this increased expressiveness, the learner is able to
    represent uncertainty and a lack of knowledge in a more flexible and more faithful
    manner. To learn from weakly labeled data of that kind, we leverage methods that
    have recently been proposed in the realm of so-called superset learning. In an
    exhaustive empirical evaluation, we compare our methodology to state-of-the-art
    self-supervision approaches, showing competitive to superior performance especially
    in low-label scenarios incorporating a high degree of uncertainty.
author:
- first_name: Julian
  full_name: Lienen, Julian
  id: '44040'
  last_name: Lienen
- first_name: Eyke
  full_name: Hüllermeier, Eyke
  id: '48129'
  last_name: Hüllermeier
citation:
  ama: Lienen J, Hüllermeier E. Credal Self-Supervised Learning. <i>arXiv:210611853</i>.
    2021.
  apa: Lienen, J., &#38; Hüllermeier, E. (2021). Credal Self-Supervised Learning.
    <i>ArXiv:2106.11853</i>.
  bibtex: '@article{Lienen_Hüllermeier_2021, title={Credal Self-Supervised Learning},
    journal={arXiv:2106.11853}, author={Lienen, Julian and Hüllermeier, Eyke}, year={2021}
    }'
  chicago: Lienen, Julian, and Eyke Hüllermeier. “Credal Self-Supervised Learning.”
    <i>ArXiv:2106.11853</i>, 2021.
  ieee: J. Lienen and E. Hüllermeier, “Credal Self-Supervised Learning,” <i>arXiv:2106.11853</i>.
    2021.
  mla: Lienen, Julian, and Eyke Hüllermeier. “Credal Self-Supervised Learning.” <i>ArXiv:2106.11853</i>,
    2021.
  short: J. Lienen, E. Hüllermeier, ArXiv:2106.11853 (2021).
date_created: 2021-06-23T07:24:38Z
date_updated: 2022-01-06T06:55:35Z
language:
- iso: eng
main_file_link:
- open_access: '1'
  url: https://arxiv.org/pdf/2106.11853.pdf
oa: '1'
publication: arXiv:2106.11853
status: public
title: Credal Self-Supervised Learning
type: preprint
user_id: '44040'
year: '2021'
...
---
_id: '20211'
abstract:
- lang: eng
  text: "In many real-world applications, the relative depth of objects in an image
    is\r\ncrucial for scene understanding, e.g., to calculate occlusions in augmented\r\nreality
    scenes. Predicting depth in monocular images has recently been tackled\r\nusing
    machine learning methods, mainly by treating the problem as a regression\r\ntask.
    Yet, being interested in an order relation in the first place,\r\nranking methods
    suggest themselves as a natural alternative to regression, and\r\nindeed, ranking
    approaches leveraging pairwise comparisons as training\r\ninformation (\"object
    A is closer to the camera than B\") have shown promising\r\nperformance on this
    problem. In this paper, we elaborate on the use of\r\nso-called \\emph{listwise}
    ranking as a generalization of the pairwise approach.\r\nListwise ranking goes
    beyond pairwise comparisons between objects and considers\r\nrankings of arbitrary
    length as training information. Our approach is based on\r\nthe Plackett-Luce
    model, a probability distribution on rankings, which we\r\ncombine with a state-of-the-art
    neural network architecture and a sampling\r\nstrategy to reduce training complexity.
    An empirical evaluation on benchmark\r\ndata in a \"zero-shot\" setting demonstrates
    the effectiveness of our proposal\r\ncompared to existing ranking and regression
    methods."
author:
- first_name: Julian
  full_name: Lienen, Julian
  id: '44040'
  last_name: Lienen
- first_name: Eyke
  full_name: Hüllermeier, Eyke
  id: '48129'
  last_name: Hüllermeier
citation:
  ama: Lienen J, Hüllermeier E. Monocular Depth Estimation via Listwise Ranking using
    the Plackett-Luce  model. <i>arXiv:201013118</i>. 2020.
  apa: Lienen, J., &#38; Hüllermeier, E. (2020). Monocular Depth Estimation via Listwise
    Ranking using the Plackett-Luce  model. <i>ArXiv:2010.13118</i>.
  bibtex: '@article{Lienen_Hüllermeier_2020, title={Monocular Depth Estimation via
    Listwise Ranking using the Plackett-Luce  model}, journal={arXiv:2010.13118},
    author={Lienen, Julian and Hüllermeier, Eyke}, year={2020} }'
  chicago: Lienen, Julian, and Eyke Hüllermeier. “Monocular Depth Estimation via Listwise
    Ranking Using the Plackett-Luce  Model.” <i>ArXiv:2010.13118</i>, 2020.
  ieee: J. Lienen and E. Hüllermeier, “Monocular Depth Estimation via Listwise Ranking
    using the Plackett-Luce  model,” <i>arXiv:2010.13118</i>. 2020.
  mla: Lienen, Julian, and Eyke Hüllermeier. “Monocular Depth Estimation via Listwise
    Ranking Using the Plackett-Luce  Model.” <i>ArXiv:2010.13118</i>, 2020.
  short: J. Lienen, E. Hüllermeier, ArXiv:2010.13118 (2020).
date_created: 2020-10-27T07:48:40Z
date_updated: 2022-01-06T06:54:23Z
language:
- iso: eng
main_file_link:
- open_access: '1'
  url: https://arxiv.org/abs/2010.13118
oa: '1'
publication: arXiv:2010.13118
status: public
title: Monocular Depth Estimation via Listwise Ranking using the Plackett-Luce  model
type: preprint
user_id: '44040'
year: '2020'
...
---
_id: '16415'
author:
- first_name: Julian
  full_name: Lienen, Julian
  id: '44040'
  last_name: Lienen
citation:
  ama: Lienen J. <i>Automated Feature Engineering on Time Series Data</i>.; 2019.
  apa: Lienen, J. (2019). <i>Automated Feature Engineering on Time Series Data</i>.
  bibtex: '@book{Lienen_2019, title={Automated Feature Engineering on Time Series
    Data}, author={Lienen, Julian}, year={2019} }'
  chicago: Lienen, Julian. <i>Automated Feature Engineering on Time Series Data</i>,
    2019.
  ieee: J. Lienen, <i>Automated Feature Engineering on Time Series Data</i>. 2019.
  mla: Lienen, Julian. <i>Automated Feature Engineering on Time Series Data</i>. 2019.
  short: J. Lienen, Automated Feature Engineering on Time Series Data, 2019.
date_created: 2020-04-06T11:44:17Z
date_updated: 2022-01-06T06:52:50Z
language:
- iso: eng
status: public
title: Automated Feature Engineering on Time Series Data
type: mastersthesis
user_id: '44040'
year: '2019'
...