Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor.author | Sick, Beate | - |
| dc.contributor.author | Hathorn, Torsten | - |
| dc.contributor.author | Durr, Oliver | - |
| dc.date.accessioned | 2024-03-15T13:00:42Z | - |
| dc.date.available | 2024-03-15T13:00:42Z | - |
| dc.date.issued | 2021 | - |
| dc.identifier.isbn | 978-1-7281-8808-9 | de_CH |
| dc.identifier.uri | https://arxiv.org/pdf/2004.00464.pdf | de_CH |
| dc.identifier.uri | https://digitalcollection.zhaw.ch/handle/11475/30200 | - |
| dc.description.abstract | We present a deep transformation model for probabilistic regression. Deep learning is known for outstandingly accurate predictions on complex data but in regression tasks it is predominantly used to just predict a single number. This ignores the non-deterministic character of most tasks. Especially if crucial decisions are based on the predictions, like in medical applications, it is essential to quantify the prediction uncertainty. The presented deep learning transformation model estimates the whole conditional probability distribution, which is the most thorough way to capture uncertainty about the outcome. We combine ideas from a statistical transformation model (most likely transformation) with recent transformation models from deep learning (normalizing flows) to predict complex outcome distributions. The core of the method is a parameterized transformation function which can be trained with the usual maximum likelihood framework using gradient descent. The method can be combined with existing deep learning architectures. For small machine learning benchmark datasets, we report state of the art performance for most dataset and partly even outperform it. Our method works for complex input data, which we demonstrate by employing a CNN architecture on image data. | de_CH |
| dc.language.iso | en | de_CH |
| dc.publisher | IEEE | de_CH |
| dc.rights | Licence according to publishing contract | de_CH |
| dc.subject | Deep learning | de_CH |
| dc.subject | Maximum likelihood estimation | de_CH |
| dc.subject | Uncertainty | de_CH |
| dc.subject | Neural networks | de_CH |
| dc.subject | Medical service | de_CH |
| dc.subject | Predictive model | de_CH |
| dc.subject | Probabilistic logic | de_CH |
| dc.subject.ddc | 006: Spezielle Computerverfahren | de_CH |
| dc.subject.ddc | 510: Mathematik | de_CH |
| dc.title | Deep transformation models : tackling complex regression problems with neural network based transformation models | de_CH |
| dc.type | Konferenz: Paper | de_CH |
| dcterms.type | Text | de_CH |
| zhaw.departement | School of Engineering | de_CH |
| zhaw.organisationalunit | Institut für Datenanalyse und Prozessdesign (IDP) | de_CH |
| dc.identifier.doi | 10.1109/ICPR48806.2021.9413177 | de_CH |
| zhaw.conference.details | 25th International Conference on Pattern Recognition (ICPR), virtual, 10-15 January 2021 | de_CH |
| zhaw.funding.eu | No | de_CH |
| zhaw.originated.zhaw | Yes | de_CH |
| zhaw.pages.end | 2481 | de_CH |
| zhaw.pages.start | 2476 | de_CH |
| zhaw.publication.status | publishedVersion | de_CH |
| zhaw.publication.review | Peer review (Publikation) | de_CH |
| zhaw.title.proceedings | 2020 25th International Conference on Pattern Recognition (ICPR) | de_CH |
| zhaw.author.additional | No | de_CH |
| zhaw.display.portrait | Yes | de_CH |
| Appears in collections: | Publikationen School of Engineering | |
Files in This Item:
There are no files associated with this item.
Show simple item record
Sick, B., Hathorn, T., & Durr, O. (2021). Deep transformation models : tackling complex regression problems with neural network based transformation models [Conference paper]. 2020 25th International Conference on Pattern Recognition (ICPR), 2476–2481. https://doi.org/10.1109/ICPR48806.2021.9413177
Sick, B., Hathorn, T. and Durr, O. (2021) ‘Deep transformation models : tackling complex regression problems with neural network based transformation models’, in 2020 25th International Conference on Pattern Recognition (ICPR). IEEE, pp. 2476–2481. Available at: https://doi.org/10.1109/ICPR48806.2021.9413177.
B. Sick, T. Hathorn, and O. Durr, “Deep transformation models : tackling complex regression problems with neural network based transformation models,” in 2020 25th International Conference on Pattern Recognition (ICPR), 2021, pp. 2476–2481. doi: 10.1109/ICPR48806.2021.9413177.
SICK, Beate, Torsten HATHORN und Oliver DURR, 2021. Deep transformation models : tackling complex regression problems with neural network based transformation models. In: 2020 25th International Conference on Pattern Recognition (ICPR) [online]. Conference paper. IEEE. 2021. S. 2476–2481. ISBN 978-1-7281-8808-9. Verfügbar unter: https://arxiv.org/pdf/2004.00464.pdf
Sick, Beate, Torsten Hathorn, and Oliver Durr. 2021. “Deep Transformation Models : Tackling Complex Regression Problems with Neural Network Based Transformation Models.” Conference paper. In 2020 25th International Conference on Pattern Recognition (ICPR), 2476–81. IEEE. https://doi.org/10.1109/ICPR48806.2021.9413177.
Sick, Beate, et al. “Deep Transformation Models : Tackling Complex Regression Problems with Neural Network Based Transformation Models.” 2020 25th International Conference on Pattern Recognition (ICPR), IEEE, 2021, pp. 2476–81, https://doi.org/10.1109/ICPR48806.2021.9413177.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.