---
OA_place: repository
OA_type: green
_id: '22294'
abstract:
- lang: eng
text: 'Modern computer systems store vast amounts of personal data, enabling advances
in AI and ML but risking user privacy and trust. For privacy reasons, it is sometimes
desired for an ML model to forget part of the data it was trained on. In this
paper, we introduce a novel unlearning approach based on Forgetting Neural Networks
(FNNs), a neuroscience-inspired architecture that explicitly encodes forgetting
through multiplicative decay factors. While FNNs had previously been studied as
a theoretical construct, we provide the first concrete implementation and demonstrate
their effectiveness for targeted unlearning. We propose several variants with
per-neuron forgetting factors, including rank-based assignments guided by activation
levels, and evaluate them on MNIST and Fashion-MNIST benchmarks. Our method systematically
removes information associated with forget sets while preserving performance on
retained data. Membership inference attacks confirm the effectiveness of FNN-based
unlearning in erasing information about the training data from the neural network.
These results establish FNNs as a promising foundation for efficient and interpretable
unlearning. '
article_processing_charge: No
arxiv: 1
author:
- first_name: Amartya
full_name: Hatua, Amartya
last_name: Hatua
- first_name: Trung
full_name: Nguyen, Trung
last_name: Nguyen
- first_name: Filip
full_name: Cano Cordoba, Filip
id: 708cad98-e86a-11ef-8098-bdae2d7c6af1
last_name: Cano Cordoba
orcid: 0000-0002-0783-904X
- first_name: Andrew
full_name: Sung, Andrew
last_name: Sung
citation:
ama: 'Hatua A, Nguyen T, Cano Cordoba F, Sung A. Machine unlearning using forgetting
neural networks. In: Proceedings of the 18th International Conference on Agents
and Artificial Intelligence. Vol 2. SciTePress; 2026:1536-1546. doi:10.5220/0014326500004052'
apa: 'Hatua, A., Nguyen, T., Cano Cordoba, F., & Sung, A. (2026). Machine unlearning
using forgetting neural networks. In Proceedings of the 18th International
Conference on Agents and Artificial Intelligence (Vol. 2, pp. 1536–1546).
Marbella, Spain: SciTePress. https://doi.org/10.5220/0014326500004052'
chicago: Hatua, Amartya, Trung Nguyen, Filip Cano Cordoba, and Andrew Sung. “Machine
Unlearning Using Forgetting Neural Networks.” In Proceedings of the 18th International
Conference on Agents and Artificial Intelligence, 2:1536–46. SciTePress, 2026.
https://doi.org/10.5220/0014326500004052.
ieee: A. Hatua, T. Nguyen, F. Cano Cordoba, and A. Sung, “Machine unlearning using
forgetting neural networks,” in Proceedings of the 18th International Conference
on Agents and Artificial Intelligence, Marbella, Spain, 2026, vol. 2, pp.
1536–1546.
ista: 'Hatua A, Nguyen T, Cano Cordoba F, Sung A. 2026. Machine unlearning using
forgetting neural networks. Proceedings of the 18th International Conference on
Agents and Artificial Intelligence. ICAART: International Conference on Agents
and Artificial Intelligence vol. 2, 1536–1546.'
mla: Hatua, Amartya, et al. “Machine Unlearning Using Forgetting Neural Networks.”
Proceedings of the 18th International Conference on Agents and Artificial Intelligence,
vol. 2, SciTePress, 2026, pp. 1536–46, doi:10.5220/0014326500004052.
short: A. Hatua, T. Nguyen, F. Cano Cordoba, A. Sung, in:, Proceedings of the 18th
International Conference on Agents and Artificial Intelligence, SciTePress, 2026,
pp. 1536–1546.
conference:
end_date: 2026-03-08
location: Marbella, Spain
name: 'ICAART: International Conference on Agents and Artificial Intelligence'
start_date: 2026-03-05
das_tickbox: '1'
date_created: 2026-07-13T09:46:46Z
date_published: 2026-06-30T00:00:00Z
date_updated: 2026-07-16T09:02:53Z
day: '30'
department:
- _id: ToHe
doi: 10.5220/0014326500004052
external_id:
arxiv:
- '2410.22374'
fulldoi: https://doi.org/10.5220/0014326500004052
intvolume: ' 2'
keyword:
- Machine Unlearning
- Neuroscience-Inspired Machine Learning
- Membership Inference Attacks
language:
- iso: eng
main_file_link:
- open_access: '1'
url: https://doi.org/10.48550/arXiv.2410.22374
month: '06'
oa: 1
oa_version: Preprint
page: 1536-1546
publication: Proceedings of the 18th International Conference on Agents and Artificial
Intelligence
publication_identifier:
eissn:
- 2184-433X
isbn:
- '9789897587962'
publication_status: published
publisher: SciTePress
quality_controlled: '1'
scopus_import: '1'
status: public
title: Machine unlearning using forgetting neural networks
type: conference
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
volume: 2
year: '2026'
...
---
OA_place: publisher
OA_type: gold
_id: '22146'
abstract:
- lang: eng
text: We study differentially private model training with stochastic gradient descent
under learning rate scheduling and correlated noise. Although correlated noise,
in particular via matrix factorizations, has been shown to improve accuracy, prior
theoretical work focused primarily on the prefix-sum workload. That workload assumes
a constant learning rate, whereas in practice learning rate schedules are widely
used to accelerate training and improve convergence. We close this gap by deriving
general upper and lower bounds for a broad class of learning rate schedules in
both single- and multi-epoch settings. Building on these results, we propose a
learning-rate-aware factorization that achieves improvements over prefix-sum factorizations
under both MaxSE and MeanSE error metrics. Our theoretical analysis yields memory-efficient
constructions suitable for practical deployment, and experiments on CIFAR-10 and
IMDB datasets confirm that schedule-aware factorizations improve accuracy in private
training.
acknowledgement: "We thank Rasmus Pagh, Christoph Lampert and Jalaj Upadhyay for valuable\r\ncomments
on an early draft. We thank Ryan Mckenna for a fruitful discussion on the experiment\r\ndesign.
We thank Antti Honkela for sharing insights on learning rate scheduling and DP.\r\nNikita
P. Kalinin: Funded in part by the Austrian Science Fund (FWF) [10.55776/COE12].\r\nJoel
Daniel Andersson: Funded by the European Union. Views and opinions expressed are
however\r\nthose of the author(s) only and do not necessarily reflect those of the
European Union or the European\r\nResearch Council Executive Agency. Neither the
European Union nor the granting authority can be\r\nheld responsible for them. This
project has received funding from the European Research Council\r\n(ERC) under the
European Union’s Horizon 2020 research and innovation programme (MoDynStruct,\r\nNo.
101019564). Additional funding by Providentia, a Data Science Distinguished Investigator
grant\r\nfrom Novo Nordisk Fonden, with additional support from VILLUM Investigator
grant 54451.\r\n"
alternative_title:
- LIPIcs
article_number: 2:1-2:21
article_processing_charge: No
arxiv: 1
author:
- first_name: Nikita
full_name: Kalinin, Nikita
id: 4b14526e-14d2-11ed-ba64-c14c9553d137
last_name: Kalinin
- first_name: Joel D
full_name: Andersson, Joel D
id: 4a893819-d954-11f0-89b1-e360bad9ccc5
last_name: Andersson
citation:
ama: 'Kalinin N, Andersson JD. Learning rate scheduling with matrix factorization
for private training. In: 7th Symposium on Foundations of Responsible Computing.
Vol 368. Schloss Dagstuhl - Leibniz-Zentrum für Informatik; 2026. doi:10.4230/LIPIcs.FORC.2026.2'
apa: 'Kalinin, N., & Andersson, J. D. (2026). Learning rate scheduling with
matrix factorization for private training. In 7th Symposium on Foundations
of Responsible Computing (Vol. 368). Cambridge, MA; United States: Schloss
Dagstuhl - Leibniz-Zentrum für Informatik. https://doi.org/10.4230/LIPIcs.FORC.2026.2'
chicago: Kalinin, Nikita, and Joel D Andersson. “Learning Rate Scheduling with Matrix
Factorization for Private Training.” In 7th Symposium on Foundations of Responsible
Computing, Vol. 368. Schloss Dagstuhl - Leibniz-Zentrum für Informatik, 2026.
https://doi.org/10.4230/LIPIcs.FORC.2026.2.
ieee: N. Kalinin and J. D. Andersson, “Learning rate scheduling with matrix factorization
for private training,” in 7th Symposium on Foundations of Responsible Computing,
Cambridge, MA; United States, 2026, vol. 368.
ista: 'Kalinin N, Andersson JD. 2026. Learning rate scheduling with matrix factorization
for private training. 7th Symposium on Foundations of Responsible Computing. FORC:
Symposium on Foundations of Responsible Computing, LIPIcs, vol. 368, 2:1-2:21.'
mla: Kalinin, Nikita, and Joel D. Andersson. “Learning Rate Scheduling with Matrix
Factorization for Private Training.” 7th Symposium on Foundations of Responsible
Computing, vol. 368, 2:1-2:21, Schloss Dagstuhl - Leibniz-Zentrum für Informatik,
2026, doi:10.4230/LIPIcs.FORC.2026.2.
short: N. Kalinin, J.D. Andersson, in:, 7th Symposium on Foundations of Responsible
Computing, Schloss Dagstuhl - Leibniz-Zentrum für Informatik, 2026.
conference:
end_date: 2026-06-05
location: Cambridge, MA; United States
name: 'FORC: Symposium on Foundations of Responsible Computing'
start_date: 2026-06-03
corr_author: '1'
das_tickbox: '0'
date_created: 2026-06-28T22:01:34Z
date_published: 2026-06-01T00:00:00Z
date_updated: 2026-09-16T07:37:21Z
day: '01'
ddc:
- '000'
department:
- _id: ChLa
- _id: GradSch
- _id: MoHe
doi: 10.4230/LIPIcs.FORC.2026.2
ec_funded: 1
external_id:
arxiv:
- '2511.17994'
file:
- access_level: open_access
checksum: c661f016d3861a1c1b590b87a744d087
content_type: application/pdf
creator: dernst
date_created: 2026-06-29T06:55:23Z
date_updated: 2026-06-29T06:55:23Z
file_id: '22149'
file_name: 2026_LIPIcsFORC_Kalinin.pdf
file_size: 1231914
relation: main_file
success: 1
file_date_updated: 2026-06-29T06:55:23Z
fulldoi: https://doi.org/10.4230/LIPIcs.FORC.2026.2
has_accepted_license: '1'
intvolume: ' 368'
keyword:
- differential privacy
- machine learning
- matrix factorization
language:
- iso: eng
license: https://creativecommons.org/licenses/by/4.0/
month: '06'
oa: 1
oa_version: Published Version
project:
- _id: bd9ca328-d553-11ed-ba76-dc4f890cfe62
call_identifier: H2020
grant_number: '101019564'
name: The design and evaluation of modern fully dynamic data structures
- _id: d8f03aaa-b035-11f1-8588-d5147fa879e0
grant_number: COE12
name: Bilateral Artificial Intelligence (Lampert)
publication: 7th Symposium on Foundations of Responsible Computing
publication_identifier:
eissn:
- 1868-8969
isbn:
- '9783959774192'
publication_status: published
publisher: Schloss Dagstuhl - Leibniz-Zentrum für Informatik
quality_controlled: '1'
researchdata_availability: no
scopus_import: '1'
status: public
supplementarymaterial: no
title: Learning rate scheduling with matrix factorization for private training
tmp:
image: /images/cc_by.png
legal_code_url: https://creativecommons.org/licenses/by/4.0/legalcode
name: Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)
short: CC BY (4.0)
type: conference
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
volume: 368
year: '2026'
...
---
OA_place: publisher
_id: '22873'
acknowledged_ssus:
- _id: ScienComp
- _id: CampIT
acknowledgement: I acknowledge funding by the Austrian Science Fund (FWF) [10.55776/PAT8537123].
alternative_title:
- ISTA Thesis
article_processing_charge: No
author:
- first_name: Andreas
full_name: Ehrmann, Andreas
id: eaa689ed-f6e0-11ea-865d-bd98cbcf83c2
last_name: Ehrmann
orcid: 0000-0002-0997-5678
citation:
ama: 'Ehrmann A. Biological functionality without biochemistry: Designing nanomachines
for target behavior. 2026. doi:10.15479/AT-ISTA-22873'
apa: 'Ehrmann, A. (2026). Biological functionality without biochemistry: Designing
nanomachines for target behavior. Institute of Science and Technology Austria.
https://doi.org/10.15479/AT-ISTA-22873'
chicago: 'Ehrmann, Andreas. “Biological Functionality without Biochemistry: Designing
Nanomachines for Target Behavior.” Institute of Science and Technology Austria,
2026. https://doi.org/10.15479/AT-ISTA-22873.'
ieee: 'A. Ehrmann, “Biological functionality without biochemistry: Designing nanomachines
for target behavior,” Institute of Science and Technology Austria, 2026.'
ista: 'Ehrmann A. 2026. Biological functionality without biochemistry: Designing
nanomachines for target behavior. Institute of Science and Technology Austria.'
mla: 'Ehrmann, Andreas. Biological Functionality without Biochemistry: Designing
Nanomachines for Target Behavior. Institute of Science and Technology Austria,
2026, doi:10.15479/AT-ISTA-22873.'
short: 'A. Ehrmann, Biological Functionality without Biochemistry: Designing Nanomachines
for Target Behavior, Institute of Science and Technology Austria, 2026.'
corr_author: '1'
das_tickbox: '0'
date_created: 2026-09-09T12:18:24Z
date_published: 2026-09-04T00:00:00Z
date_updated: 2026-09-18T11:28:29Z
day: '04'
ddc:
- '530'
- '600'
- '621'
- '004'
- '005'
degree_awarded: PhD
department:
- _id: GradSch
- _id: CaGo
- _id: EdHa
doi: 10.15479/AT-ISTA-22873
doi_confirm: '1'
file:
- access_level: closed
checksum: a5e3d79e0e5f3f4fb3c414e48116dc9e
content_type: application/pdf
creator: aehrmann
date_created: 2026-09-11T07:23:57Z
date_updated: 2026-09-17T08:59:39Z
embargo: 2027-01-15
embargo_to: open_access
file_id: '22901'
file_name: 2026_Ehrmann_Andreas_Thesis.pdf
file_size: 14782662
relation: main_file
- access_level: closed
checksum: 0235b9c52fee0c97f402f1c3d4efc1eb
content_type: application/zip
creator: aehrmann
date_created: 2026-09-11T07:24:14Z
date_updated: 2026-09-11T07:24:14Z
description: All LaTeX files to compile my PhD Thesis.
file_id: '22902'
file_name: thesis_latex_files.zip
file_size: 155228
relation: source_file
file_date_updated: 2026-09-17T08:59:39Z
fulldoi: https://doi.org/10.15479/AT-ISTA-22873
has_accepted_license: '1'
keyword:
- PhD Thesis
- functional nanomachines
- biological functionality
- nanotechnology
- energy delivery
- target behavior
- dynamics
- design principles
- optimization
- differentiable statistical physics
- machine learning
language:
- iso: eng
license: https://creativecommons.org/licenses/by-nc-sa/4.0/
month: '09'
oa_version: Published Version
page: '168'
project:
- _id: 90a98bb5-16d5-11f0-9cad-9675f3f8015d
grant_number: PAT 8537123
name: Functional bio-inspired nanomachines from sticky colloids
publication_identifier:
isbn:
- 978-3-99078-092-3
issn:
- 2663-337X
publication_status: published
publisher: Institute of Science and Technology Austria
related_material:
record:
- id: '22893'
relation: part_of_dissertation
status: public
- id: '22892'
relation: part_of_dissertation
status: public
researchdata_availability: upon request
status: public
supervisor:
- first_name: Carl Peter
full_name: Goodrich, Carl Peter
id: EB352CD2-F68A-11E9-89C5-A432E6697425
last_name: Goodrich
orcid: 0000-0002-1307-5074
- first_name: Edouard B
full_name: Hannezo, Edouard B
id: 3A9DB764-F248-11E8-B48F-1D18A9856A87
last_name: Hannezo
orcid: 0000-0001-6005-1561
supplementarymaterial: not applicable
title: 'Biological functionality without biochemistry: Designing nanomachines for
target behavior'
tmp:
image: /images/cc_by_nc_sa.png
legal_code_url: https://creativecommons.org/licenses/by-nc-sa/4.0/legalcode
name: Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC
BY-NC-SA 4.0)
short: CC BY-NC-SA (4.0)
type: dissertation
user_id: 8b945eb4-e2f2-11eb-945a-df72226e66a9
year: '2026'
...
---
OA_place: publisher
_id: '22857'
abstract:
- lang: eng
text: "Artificial intelligence and machine learning have undergone an unprecedented
evolution in the past decade, motivating a research effort toward a theory able
to capture the qualitative behavior of large-scale neural systems. A central puzzle
has been the clear benefit of scaling architecture size and overfitting the training
set in supervised learning tasks. This evidence, in apparent contradiction with
classical statistical learning theory, pushed researchers to develop a new theory
capturing the interplay between the algorithmic and architectural bias of training
and the specific target function, differently from previous methods rooted in
uniform stability.\r\nThis approach has enabled a grounded understanding of novel
learning regimes, typically through formal limits where the number of training
samples $n$, data dimensions $d$, and model parameters $p$ grow to infinity at
different rates. \\\\\r\nIn this thesis, we follow this approach, focusing on
the trustworthiness of high-dimensional models: properties that are difficult
to control during training or deployment and often emerge under unpredictable
or adversarial conditions. In such settings, it is crucial to formally ensure
a priori the reliability of machine learning systems.\r\nFirst, we study data
memorization, both as label fitting and as the storage of private information
about training samples in trained parameters. We prove that $p = \\Omega(n)$ parameters
are sufficient for a deep neural network to memorize a generic set of labels,
and for a model to memorize spurious features across training data. We then give
evidence that $p = \\Omega(dn)$ parameters are instead necessary for an adversary
to reconstruct the full training set from the trained parameters.\r\nSecond, we
study robustness, both to adversarial perturbations and to distribution shift.
We first prove that $p = \\Omega(dn)$ parameters can be sufficient for a class
of neural networks to overfit the training data while guaranteeing robustness
to adversarial perturbations. Then, we focus on spurious correlations learning
in high-dimensional regression, studying the effect of the ridge regularization
parameter in the proportional regime $n = \\Theta(d)$, and connecting it via an
equivalence argument to the role of over-parameterization $p = \\Omega(n)$ in
neural networks. We also investigate the architectural bias of attention-based
networks, showing that they are sensitive to the replacement of individual words
in an embedded sentence, allowing them to generalize on sentences where the contextual
meaning depends on one or few words.\r\nFinally, we study differentially private
optimization in high-dimensional regimes. We prove that standard private gradient
methods do not suffer in the over-parameterized regime $p = \\Omega(n)$, challenging
the current wisdom based on stability-derived generalization bounds. We then consider
linear regression in the proportional regime $n = \\Theta(d)$, showing that standard
private gradient descent can achieve optimal rates under appropriate hyper-parameter
scaling, such as sufficiently small gradient clipping constants, whose role is
still debated in practice."
acknowledged_ssus:
- _id: ScienComp
acknowledgement: "This project was partially supported by the 2019 Lopez-Loreta prize,\r\nthe
European Union (ERC, INF2\r\n, project number 101161364), the Austrian Science Fund\r\n(FWF)
10.55776/COE12, and a Google PhD fellowship in machine intelligence. Furthermore,\r\nthe
candidate acknowledges the support from the Scientific Service Units of the Institute
of\r\nScience and Technology Austria through resources provided by Scientific Computing."
alternative_title:
- ISTA Thesis
article_processing_charge: No
author:
- first_name: Simone
full_name: Bombari, Simone
id: ca726dda-de17-11ea-bc14-f9da834f63aa
last_name: Bombari
citation:
ama: Bombari S. Trustworthy machine learning in high dimensions. 2026. doi:10.15479/AT-ISTA-22857
apa: Bombari, S. (2026). Trustworthy machine learning in high dimensions.
Institute of Science and Technology Austria. https://doi.org/10.15479/AT-ISTA-22857
chicago: Bombari, Simone. “Trustworthy Machine Learning in High Dimensions.” Institute
of Science and Technology Austria, 2026. https://doi.org/10.15479/AT-ISTA-22857.
ieee: S. Bombari, “Trustworthy machine learning in high dimensions,” Institute of
Science and Technology Austria, 2026.
ista: Bombari S. 2026. Trustworthy machine learning in high dimensions. Institute
of Science and Technology Austria.
mla: Bombari, Simone. Trustworthy Machine Learning in High Dimensions. Institute
of Science and Technology Austria, 2026, doi:10.15479/AT-ISTA-22857.
short: S. Bombari, Trustworthy Machine Learning in High Dimensions, Institute of
Science and Technology Austria, 2026.
corr_author: '1'
das_tickbox: '0'
date_created: 2026-09-08T13:40:08Z
date_published: 2026-09-08T00:00:00Z
date_updated: 2026-09-21T13:07:01Z
day: '08'
ddc:
- '519'
degree_awarded: PhD
department:
- _id: GradSch
- _id: MaMo
doi: 10.15479/AT-ISTA-22857
doi_confirm: '1'
file:
- access_level: closed
checksum: 8eab99d6dc6e4476826bdfc7c00fdebb
content_type: application/zip
creator: sbombari
date_created: 2026-09-08T13:28:38Z
date_updated: 2026-09-08T13:28:38Z
file_id: '22859'
file_name: Thesis copy.zip
file_size: 99154325
relation: source_file
- access_level: open_access
checksum: 007033bafe4622ff4c2e758301354df1
content_type: application/pdf
creator: sbombari
date_created: 2026-09-10T10:03:48Z
date_updated: 2026-09-10T10:03:48Z
file_id: '22898'
file_name: 2026_Bombari_Simone_Thesis.pdf
file_size: 12856211
relation: main_file
success: 1
file_date_updated: 2026-09-10T10:03:48Z
fulldoi: https://doi.org/10.15479/AT-ISTA-22857
has_accepted_license: '1'
keyword:
- machine learning
- high-dimensional statistics
- deep learning theory
- privacy
- memorization
- robustness
language:
- iso: eng
month: '09'
oa: 1
oa_version: Published Version
page: '446'
project:
- _id: 92099302-16d5-11f0-9cad-f9a785f54fbd
name: 'Trustworthy Deep Learning Theory: Private Over-Parameterized Models and Robust
LLMs'
- _id: 911e6d1f-16d5-11f0-9cad-c5c68c6a1cdf
grant_number: '101161364'
name: 'Inference in High Dimensions: Light-speed Algorithms and Information Limits'
- _id: 74caaef7-b034-11f1-8f2d-e0e993bb422e
grant_number: COE12
name: Bilateral Artificial Intelligence (Mondelli)
publication_identifier:
isbn:
- 978-3-99078-091-6
issn:
- 2663-337X
publication_status: published
publisher: Institute of Science and Technology Austria
related_material:
record:
- id: '12537'
relation: part_of_dissertation
status: public
- id: '18972'
relation: part_of_dissertation
status: public
- id: '18973'
relation: part_of_dissertation
status: public
- id: '21324'
relation: part_of_dissertation
status: public
- id: '22894'
relation: part_of_dissertation
status: public
- id: '19627'
relation: part_of_dissertation
status: public
- id: '12859'
relation: part_of_dissertation
status: public
researchdata_availability: no
status: public
supervisor:
- first_name: Marco
full_name: Mondelli, Marco
id: 27EB676C-8706-11E9-9510-7717E6697425
last_name: Mondelli
orcid: 0000-0002-3242-7020
supplementarymaterial: no
title: Trustworthy machine learning in high dimensions
type: dissertation
user_id: 8b945eb4-e2f2-11eb-945a-df72226e66a9
year: '2026'
...
---
OA_place: publisher
_id: '10799'
abstract:
- lang: eng
text: "Because of the increasing popularity of machine learning methods, it is becoming
important to understand the impact of learned components on automated decision-making
systems and to guarantee that their consequences are beneficial to society. In
other words, it is necessary to ensure that machine learning is sufficiently trustworthy
to be used in real-world applications. This thesis studies two properties of machine
learning models that are highly desirable for the\r\nsake of reliability: robustness
and fairness. In the first part of the thesis we study the robustness of learning
algorithms to training data corruption. Previous work has shown that machine learning
models are vulnerable to a range\r\nof training set issues, varying from label
noise through systematic biases to worst-case data manipulations. This is an especially
relevant problem from a present perspective, since modern machine learning methods
are particularly data hungry and therefore practitioners often have to rely on
data collected from various external sources, e.g. from the Internet, from app
users or via crowdsourcing. Naturally, such sources vary greatly in the quality
and reliability of the\r\ndata they provide. With these considerations in mind,
we study the problem of designing machine learning algorithms that are robust
to corruptions in data coming from multiple sources. We show that, in contrast
to the case of a single dataset with outliers, successful learning within this
model is possible both theoretically and practically, even under worst-case data
corruptions. The second part of this thesis deals with fairness-aware machine
learning. There are multiple areas where machine learning models have shown promising
results, but where careful considerations are required, in order to avoid discrimanative
decisions taken by such learned components. Ensuring fairness can be particularly
challenging, because real-world training datasets are expected to contain various
forms of historical bias that may affect the learning process. In this thesis
we show that data corruption can indeed render the problem of achieving fairness
impossible, by tightly characterizing the theoretical limits of fair learning
under worst-case data manipulations. However, assuming access to clean data, we
also show how fairness-aware learning can be made practical in contexts beyond
binary classification, in particular in the challenging learning to rank setting."
alternative_title:
- ISTA Thesis
article_processing_charge: No
author:
- first_name: Nikola H
full_name: Konstantinov, Nikola H
id: 4B9D76E4-F248-11E8-B48F-1D18A9856A87
last_name: Konstantinov
orcid: 0009-0009-5204-7621
citation:
ama: Konstantinov NH. Robustness and fairness in machine learning. 2022. doi:10.15479/at:ista:10799
apa: Konstantinov, N. H. (2022). Robustness and fairness in machine learning.
Institute of Science and Technology Austria. https://doi.org/10.15479/at:ista:10799
chicago: Konstantinov, Nikola H. “Robustness and Fairness in Machine Learning.”
Institute of Science and Technology Austria, 2022. https://doi.org/10.15479/at:ista:10799.
ieee: N. H. Konstantinov, “Robustness and fairness in machine learning,” Institute
of Science and Technology Austria, 2022.
ista: Konstantinov NH. 2022. Robustness and fairness in machine learning. Institute
of Science and Technology Austria.
mla: Konstantinov, Nikola H. Robustness and Fairness in Machine Learning.
Institute of Science and Technology Austria, 2022, doi:10.15479/at:ista:10799.
short: N.H. Konstantinov, Robustness and Fairness in Machine Learning, Institute
of Science and Technology Austria, 2022.
corr_author: '1'
date_created: 2022-02-28T13:03:49Z
date_published: 2022-03-08T00:00:00Z
date_updated: 2026-04-07T14:19:48Z
day: '08'
ddc:
- '000'
degree_awarded: PhD
department:
- _id: GradSch
- _id: ChLa
doi: 10.15479/at:ista:10799
ec_funded: 1
file:
- access_level: open_access
checksum: 626bc523ae8822d20e635d0e2d95182e
content_type: application/pdf
creator: nkonstan
date_created: 2022-03-06T11:42:54Z
date_updated: 2022-03-06T11:42:54Z
file_id: '10823'
file_name: thesis.pdf
file_size: 4204905
relation: main_file
success: 1
- access_level: closed
checksum: e2ca2b88350ac8ea1515b948885cbcb1
content_type: application/x-zip-compressed
creator: nkonstan
date_created: 2022-03-06T11:42:57Z
date_updated: 2022-03-10T12:11:48Z
file_id: '10824'
file_name: thesis.zip
file_size: 22841103
relation: source_file
file_date_updated: 2022-03-10T12:11:48Z
fulldoi: https://doi.org/10.15479/at:ista:10799
has_accepted_license: '1'
keyword:
- robustness
- fairness
- machine learning
- PAC learning
- adversarial learning
language:
- iso: eng
month: '03'
oa: 1
oa_version: Published Version
page: '176'
project:
- _id: 2564DBCA-B435-11E9-9278-68D0E5697425
call_identifier: H2020
grant_number: '665385'
name: International IST Doctoral Program
publication_identifier:
isbn:
- 978-3-99078-015-2
issn:
- 2663-337X
publication_status: published
publisher: Institute of Science and Technology Austria
related_material:
record:
- id: '10802'
relation: part_of_dissertation
status: public
- id: '10803'
relation: part_of_dissertation
status: public
- id: '6590'
relation: part_of_dissertation
status: public
- id: '8724'
relation: part_of_dissertation
status: public
status: public
supervisor:
- first_name: Christoph
full_name: Lampert, Christoph
id: 40C20FD2-F248-11E8-B48F-1D18A9856A87
last_name: Lampert
orcid: 0000-0001-8622-7887
title: Robustness and fairness in machine learning
type: dissertation
user_id: ba8df636-2132-11f1-aed0-ed93e2281fdd
year: '2022'
...
---
_id: '10802'
abstract:
- lang: eng
text: "Addressing fairness concerns about machine learning models is a crucial step
towards their long-term adoption in real-world automated systems. While many approaches
have been developed for training fair models from data, little is known about
the robustness of these methods to data corruption. In this work we consider fairness-aware
learning under worst-case data manipulations. We show that an adversary can in
some situations force any learner to return an overly biased classifier, regardless
of the sample size and with or without degrading\r\naccuracy, and that the strength
of the excess bias increases for learning problems with underrepresented protected
groups in the data. We also prove that our hardness results are tight up to constant
factors. To this end, we study two natural learning algorithms that optimize for
both accuracy and fairness and show that these algorithms enjoy guarantees that
are order-optimal in terms of the corruption ratio and the protected groups frequencies
in the large data\r\nlimit."
acknowledgement: The authors thank Eugenia Iofinova and Bernd Prach for providing
feedback on early versions of this paper. This publication was made possible by
an ETH AI Center postdoctoral fellowship to Nikola Konstantinov.
article_processing_charge: No
article_type: original
arxiv: 1
author:
- first_name: Nikola H
full_name: Konstantinov, Nikola H
id: 4B9D76E4-F248-11E8-B48F-1D18A9856A87
last_name: Konstantinov
orcid: 0009-0009-5204-7621
- first_name: Christoph
full_name: Lampert, Christoph
id: 40C20FD2-F248-11E8-B48F-1D18A9856A87
last_name: Lampert
orcid: 0000-0002-4561-241X
citation:
ama: Konstantinov NH, Lampert C. Fairness-aware PAC learning from corrupted data.
Journal of Machine Learning Research. 2022;23:1-60.
apa: Konstantinov, N. H., & Lampert, C. (2022). Fairness-aware PAC learning
from corrupted data. Journal of Machine Learning Research. ML Research
Press.
chicago: Konstantinov, Nikola H, and Christoph Lampert. “Fairness-Aware PAC Learning
from Corrupted Data.” Journal of Machine Learning Research. ML Research
Press, 2022.
ieee: N. H. Konstantinov and C. Lampert, “Fairness-aware PAC learning from corrupted
data,” Journal of Machine Learning Research, vol. 23. ML Research Press,
pp. 1–60, 2022.
ista: Konstantinov NH, Lampert C. 2022. Fairness-aware PAC learning from corrupted
data. Journal of Machine Learning Research. 23, 1–60.
mla: Konstantinov, Nikola H., and Christoph Lampert. “Fairness-Aware PAC Learning
from Corrupted Data.” Journal of Machine Learning Research, vol. 23, ML
Research Press, 2022, pp. 1–60.
short: N.H. Konstantinov, C. Lampert, Journal of Machine Learning Research 23 (2022)
1–60.
corr_author: '1'
date_created: 2022-02-28T14:05:42Z
date_published: 2022-05-01T00:00:00Z
date_updated: 2026-04-07T14:19:48Z
day: '01'
ddc:
- '004'
department:
- _id: ChLa
external_id:
arxiv:
- '2102.06004'
file:
- access_level: open_access
checksum: 9cac897b54a0ddf3a553a2c33e88cfda
content_type: application/pdf
creator: kschuh
date_created: 2022-07-12T15:08:28Z
date_updated: 2022-07-12T15:08:28Z
file_id: '11570'
file_name: 2022_JournalMachineLearningResearch_Konstantinov.pdf
file_size: 551862
relation: main_file
success: 1
file_date_updated: 2022-07-12T15:08:28Z
has_accepted_license: '1'
intvolume: ' 23'
keyword:
- Fairness
- robustness
- data poisoning
- trustworthy machine learning
- PAC learning
language:
- iso: eng
month: '05'
oa: 1
oa_version: Published Version
page: 1-60
publication: Journal of Machine Learning Research
publication_identifier:
eissn:
- 1533-7928
issn:
- 1532-4435
publication_status: published
publisher: ML Research Press
quality_controlled: '1'
related_material:
record:
- id: '13241'
relation: shorter_version
status: public
- id: '10799'
relation: dissertation_contains
status: public
scopus_import: '1'
status: public
title: Fairness-aware PAC learning from corrupted data
tmp:
image: /images/cc_by.png
legal_code_url: https://creativecommons.org/licenses/by/4.0/legalcode
name: Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)
short: CC BY (4.0)
type: journal_article
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
volume: 23
year: '2022'
...
---
OA_place: publisher
_id: '11362'
abstract:
- lang: eng
text: "Deep learning has enabled breakthroughs in challenging computing problems
and has emerged as the standard problem-solving tool for computer vision and natural
language processing tasks.\r\nOne exception to this trend is safety-critical tasks
where robustness and resilience requirements contradict the black-box nature of
neural networks. \r\nTo deploy deep learning methods for these tasks, it is vital
to provide guarantees on neural network agents' safety and robustness criteria.
\r\nThis can be achieved by developing formal verification methods to verify the
safety and robustness properties of neural networks.\r\n\r\nOur goal is to design,
develop and assess safety verification methods for neural networks to improve
their reliability and trustworthiness in real-world applications.\r\nThis thesis
establishes techniques for the verification of compressed and adversarially trained
models as well as the design of novel neural networks for verifiably safe decision-making.\r\n\r\nFirst,
we establish the problem of verifying quantized neural networks. Quantization
is a technique that trades numerical precision for the computational efficiency
of running a neural network and is widely adopted in industry.\r\nWe show that
neglecting the reduced precision when verifying a neural network can lead to wrong
conclusions about the robustness and safety of the network, highlighting that
novel techniques for quantized network verification are necessary. We introduce
several bit-exact verification methods explicitly designed for quantized neural
networks and experimentally confirm on realistic networks that the network's robustness
and other formal properties are affected by the quantization.\r\n\r\nFurthermore,
we perform a case study providing evidence that adversarial training, a standard
technique for making neural networks more robust, has detrimental effects on the
network's performance. This robustness-accuracy tradeoff has been studied before
regarding the accuracy obtained on classification datasets where each data point
is independent of all other data points. On the other hand, we investigate the
tradeoff empirically in robot learning settings where a both, a high accuracy
and a high robustness, are desirable.\r\nOur results suggest that the negative
side-effects of adversarial training outweigh its robustness benefits in practice.\r\n\r\nFinally,
we consider the problem of verifying safety when running a Bayesian neural network
policy in a feedback loop with systems over the infinite time horizon. Bayesian
neural networks are probabilistic models for learning uncertainties in the data
and are therefore often used on robotic and healthcare applications where data
is inherently stochastic.\r\nWe introduce a method for recalibrating Bayesian
neural networks so that they yield probability distributions over safe decisions
only.\r\nOur method learns a safety certificate that guarantees safety over the
infinite time horizon to determine which decisions are safe in every possible
state of the system.\r\nWe demonstrate the effectiveness of our approach on a
series of reinforcement learning benchmarks."
alternative_title:
- ISTA Thesis
article_processing_charge: No
author:
- first_name: Mathias
full_name: Lechner, Mathias
id: 3DC22916-F248-11E8-B48F-1D18A9856A87
last_name: Lechner
citation:
ama: Lechner M. Learning verifiable representations. 2022. doi:10.15479/at:ista:11362
apa: Lechner, M. (2022). Learning verifiable representations. Institute of
Science and Technology Austria. https://doi.org/10.15479/at:ista:11362
chicago: Lechner, Mathias. “Learning Verifiable Representations.” Institute of Science
and Technology Austria, 2022. https://doi.org/10.15479/at:ista:11362.
ieee: M. Lechner, “Learning verifiable representations,” Institute of Science and
Technology Austria, 2022.
ista: Lechner M. 2022. Learning verifiable representations. Institute of Science
and Technology Austria.
mla: Lechner, Mathias. Learning Verifiable Representations. Institute of
Science and Technology Austria, 2022, doi:10.15479/at:ista:11362.
short: M. Lechner, Learning Verifiable Representations, Institute of Science and
Technology Austria, 2022.
corr_author: '1'
date_created: 2022-05-12T07:14:01Z
date_published: 2022-05-12T00:00:00Z
date_updated: 2026-08-19T09:28:05Z
day: '12'
ddc:
- '004'
degree_awarded: PhD
department:
- _id: GradSch
- _id: ToHe
doi: 10.15479/at:ista:11362
ec_funded: 1
file:
- access_level: closed
checksum: 8eefa9c7c10ca7e1a2ccdd731962a645
content_type: application/zip
creator: mlechner
date_created: 2022-05-13T12:33:26Z
date_updated: 2022-05-13T12:49:00Z
file_id: '11378'
file_name: src.zip
file_size: 13210143
relation: source_file
- access_level: open_access
checksum: 1b9e1e5a9a83ed9d89dad2f5133dc026
content_type: application/pdf
creator: mlechner
date_created: 2022-05-16T08:02:28Z
date_updated: 2022-05-17T15:19:39Z
file_id: '11382'
file_name: thesis_main-a2.pdf
file_size: 2732536
relation: main_file
file_date_updated: 2022-05-17T15:19:39Z
fulldoi: https://doi.org/10.15479/at:ista:11362
has_accepted_license: '1'
keyword:
- neural networks
- verification
- machine learning
language:
- iso: eng
license: https://creativecommons.org/licenses/by-nd/4.0/
month: '05'
oa: 1
oa_version: Published Version
page: '124'
project:
- _id: 25F42A32-B435-11E9-9278-68D0E5697425
call_identifier: FWF
grant_number: Z211
name: Formal methods for the design and analysis of complex systems
- _id: 62781420-2b32-11ec-9570-8d9b63373d4d
call_identifier: H2020
grant_number: '101020093'
name: Vigilant Algorithmic Monitoring of Software
publication_identifier:
isbn:
- 978-3-99078-017-6
publication_status: published
publisher: Institute of Science and Technology Austria
related_material:
record:
- id: '11366'
relation: part_of_dissertation
status: public
- id: '7808'
relation: part_of_dissertation
status: public
- id: '10666'
relation: part_of_dissertation
status: public
- id: '10667'
relation: part_of_dissertation
status: public
- id: '10665'
relation: part_of_dissertation
status: public
status: public
supervisor:
- first_name: Thomas A
full_name: Henzinger, Thomas A
id: 40876CD8-F248-11E8-B48F-1D18A9856A87
last_name: Henzinger
orcid: 0000-0002-2985-7724
title: Learning verifiable representations
tmp:
image: /image/cc_by_nd.png
legal_code_url: https://creativecommons.org/licenses/by-nd/4.0/legalcode
name: Creative Commons Attribution-NoDerivatives 4.0 International (CC BY-ND 4.0)
short: CC BY-ND (4.0)
type: dissertation
user_id: ba8df636-2132-11f1-aed0-ed93e2281fdd
year: '2022'
...
---
_id: '11627'
abstract:
- lang: eng
text: 'For a solar-like star, the surface rotation evolves with time, allowing in
principle to estimate the age of a star from its surface rotation period. Here
we are interested in measuring surface rotation periods of solar-like stars observed
by the NASA mission Kepler. Different methods have been developed to track rotation
signals in Kepler photometric light curves: time-frequency analysis based on wavelet
techniques, autocorrelation and composite spectrum. We use the learning abilities
of random forest classifiers to take decisions during two crucial steps of the
analysis. First, given some input parameters, we discriminate the considered Kepler
targets between rotating MS stars, non-rotating MS stars, red giants, binaries
and pulsators. We then use a second classifier only on the MS rotating targets
to decide the best data analysis treatment.'
article_number: '1906.09609'
article_processing_charge: No
arxiv: 1
author:
- first_name: S. N.
full_name: Breton, S. N.
last_name: Breton
- first_name: Lisa Annabelle
full_name: Bugnet, Lisa Annabelle
id: d9edb345-f866-11ec-9b37-d119b5234501
last_name: Bugnet
orcid: 0000-0003-0142-4000
- first_name: A. R. G.
full_name: Santos, A. R. G.
last_name: Santos
- first_name: A. Le
full_name: Saux, A. Le
last_name: Saux
- first_name: S.
full_name: Mathur, S.
last_name: Mathur
- first_name: P. L.
full_name: Palle, P. L.
last_name: Palle
- first_name: R. A.
full_name: Garcia, R. A.
last_name: Garcia
citation:
ama: Breton SN, Bugnet LA, Santos ARG, et al. Determining surface rotation periods
of solar-like stars observed by the Kepler mission using machine learning techniques.
arXiv. doi:10.48550/arXiv.1906.09609
apa: Breton, S. N., Bugnet, L. A., Santos, A. R. G., Saux, A. L., Mathur, S., Palle,
P. L., & Garcia, R. A. (n.d.). Determining surface rotation periods of solar-like
stars observed by the Kepler mission using machine learning techniques. arXiv.
https://doi.org/10.48550/arXiv.1906.09609
chicago: Breton, S. N., Lisa Annabelle Bugnet, A. R. G. Santos, A. Le Saux, S. Mathur,
P. L. Palle, and R. A. Garcia. “Determining Surface Rotation Periods of Solar-like
Stars Observed by the Kepler Mission Using Machine Learning Techniques.” ArXiv,
n.d. https://doi.org/10.48550/arXiv.1906.09609.
ieee: S. N. Breton et al., “Determining surface rotation periods of solar-like
stars observed by the Kepler mission using machine learning techniques,” arXiv.
.
ista: Breton SN, Bugnet LA, Santos ARG, Saux AL, Mathur S, Palle PL, Garcia RA.
Determining surface rotation periods of solar-like stars observed by the Kepler
mission using machine learning techniques. arXiv, 1906.09609.
mla: Breton, S. N., et al. “Determining Surface Rotation Periods of Solar-like Stars
Observed by the Kepler Mission Using Machine Learning Techniques.” ArXiv,
1906.09609, doi:10.48550/arXiv.1906.09609.
short: S.N. Breton, L.A. Bugnet, A.R.G. Santos, A.L. Saux, S. Mathur, P.L. Palle,
R.A. Garcia, ArXiv (n.d.).
date_created: 2022-07-20T11:18:53Z
date_published: 2019-06-23T00:00:00Z
date_updated: 2022-08-22T08:16:53Z
day: '23'
doi: 10.48550/arXiv.1906.09609
extern: '1'
external_id:
arxiv:
- '1906.09609'
fulldoi: https://doi.org/10.48550/arXiv.1906.09609
keyword:
- asteroseismology
- rotation
- solar-like stars
- kepler
- machine learning
- random forest
language:
- iso: eng
main_file_link:
- open_access: '1'
url: https://arxiv.org/abs/1906.09609
month: '06'
oa: 1
oa_version: Preprint
publication: arXiv
publication_status: submitted
status: public
title: Determining surface rotation periods of solar-like stars observed by the Kepler
mission using machine learning techniques
type: preprint
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
year: '2019'
...
---
_id: '11630'
abstract:
- lang: eng
text: 'The second mission of NASA’s Kepler satellite, K2, has collected hundreds
of thousands of lightcurves for stars close to the ecliptic plane. This new sample
could increase the number of known pulsating stars and then improve our understanding
of those stars. For the moment only a few stars have been properly classified
and published. In this work, we present a method to automaticly classify K2 pulsating
stars using a Machine Learning technique called Random Forest. The objective is
to sort out the stars in four classes: red giant (RG), main-sequence Solar-like
stars (SL), classical pulsators (PULS) and Other. To do this we use the effective
temperatures and the luminosities of the stars as well as the FliPer features,
that measures the amount of power contained in the power spectral density. The
classifier now retrieves the right classification for more than 80% of the stars.'
article_number: '1906.09611'
article_processing_charge: No
arxiv: 1
author:
- first_name: A. Le
full_name: Saux, A. Le
last_name: Saux
- first_name: Lisa Annabelle
full_name: Bugnet, Lisa Annabelle
id: d9edb345-f866-11ec-9b37-d119b5234501
last_name: Bugnet
orcid: 0000-0003-0142-4000
- first_name: S.
full_name: Mathur, S.
last_name: Mathur
- first_name: S. N.
full_name: Breton, S. N.
last_name: Breton
- first_name: R. A.
full_name: Garcia, R. A.
last_name: Garcia
citation:
ama: Saux AL, Bugnet LA, Mathur S, Breton SN, Garcia RA. Automatic classification
of K2 pulsating stars using machine learning techniques. arXiv. doi:10.48550/arXiv.1906.09611
apa: Saux, A. L., Bugnet, L. A., Mathur, S., Breton, S. N., & Garcia, R. A.
(n.d.). Automatic classification of K2 pulsating stars using machine learning
techniques. arXiv. https://doi.org/10.48550/arXiv.1906.09611
chicago: Saux, A. Le, Lisa Annabelle Bugnet, S. Mathur, S. N. Breton, and R. A.
Garcia. “Automatic Classification of K2 Pulsating Stars Using Machine Learning
Techniques.” ArXiv, n.d. https://doi.org/10.48550/arXiv.1906.09611.
ieee: A. L. Saux, L. A. Bugnet, S. Mathur, S. N. Breton, and R. A. Garcia, “Automatic
classification of K2 pulsating stars using machine learning techniques,” arXiv.
.
ista: Saux AL, Bugnet LA, Mathur S, Breton SN, Garcia RA. Automatic classification
of K2 pulsating stars using machine learning techniques. arXiv, 1906.09611.
mla: Saux, A. Le, et al. “Automatic Classification of K2 Pulsating Stars Using Machine
Learning Techniques.” ArXiv, 1906.09611, doi:10.48550/arXiv.1906.09611.
short: A.L. Saux, L.A. Bugnet, S. Mathur, S.N. Breton, R.A. Garcia, ArXiv (n.d.).
date_created: 2022-07-21T06:57:10Z
date_published: 2019-06-23T00:00:00Z
date_updated: 2022-08-22T08:20:29Z
day: '23'
doi: 10.48550/arXiv.1906.09611
extern: '1'
external_id:
arxiv:
- '1906.09611'
fulldoi: https://doi.org/10.48550/arXiv.1906.09611
keyword:
- asteroseismology - methods
- data analysis - thecniques
- machine learning - stars
- oscillations
language:
- iso: eng
main_file_link:
- open_access: '1'
url: https://doi.org/10.48550/arXiv.1906.09611
month: '06'
oa: 1
oa_version: Preprint
publication: arXiv
publication_status: submitted
status: public
title: Automatic classification of K2 pulsating stars using machine learning techniques
type: preprint
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
year: '2019'
...