---
_id: '6898'
abstract:
- lang: eng
text: 'The labeling process for supervised learning is costly and time-consuming,
and is often impractical to scale due to real-world constraints. Active learning
(AL) addresses this challenge by strategically selecting representative and informative
data points to reduce labeling efforts. This paper focuses on an AL scenario in
which only a very limited number of labels can be acquired. We propose an algorithm
operating in two phases: (1) an exploration phase that prioritizes representative
and diverse data points using density-driven criteria, and (2) an exploitation
phase that combines predictive uncertainty with density weighting to select informative
samples from densely populated regions. This enhances both representativeness
and informativeness. Our results demonstrate significant improvements in model
quality compared to other algorithms typically employed for this scenario, across
various scenarios involving imbalanced data in classification tasks and skewness
in regression tasks. Through this work, we aim to provide a new algorithm for
this scenario and investigate general principles for AL. While most AL studies
focus on either classification or regression, our work applies the algorithms
to both. Therefore, we can analyze the differences between classification and
regression problems and their effects on AL strategies. Furthermore, we explore
different categories of AL criteria and their effectiveness in the low-budget
regime. These results also provide insight into the cold-start problem, which
involves selecting an initial labeled set and is faced by many model-based AL
methods.'
alternative_id:
- '6899'
author:
- first_name: Bjarne
full_name: Jaster, Bjarne
id: '252434'
last_name: Jaster
orcid: 0000-0002-8362-5369
- first_name: Alaa
full_name: Tharwat, Alaa
id: '238549'
last_name: Tharwat
- first_name: Eiram Mahera
full_name: Sheikh, Eiram Mahera
last_name: Sheikh
- first_name: Martin
full_name: Kohlhase, Martin
id: '226669'
last_name: Kohlhase
orcid: 0009-0002-9374-0720
- first_name: Wolfram
full_name: Schenck, Wolfram
id: '224375'
last_name: Schenck
orcid: 0000-0003-3300-2048
citation:
alphadin: 'Jaster, Bjarne ; Tharwat, Alaa ; Sheikh,
Eiram Mahera ; Kohlhase, Martin
; Schenck, Wolfram: Low Query Budget
Active Learning for Classification and Regression. In: Koprinska,
I. ; Mendes-Moreira, J. ;
Branco, P. (Hrsg.): Machine Learning
and Principles and Practice of Knowledge Discovery in Databases. International
Workshops of ECML PKDD 2025, Porto, Portugal, September 15–19, 2025, Revised Selected
Papers, Part IV, Communications in Computer and Information Science.
Cham : Springer Nature Switzerland, 2026, S. 5–21'
ama: 'Jaster B, Tharwat A, Sheikh EM, Kohlhase M, Schenck W. Low Query Budget Active
Learning for Classification and Regression. In: Koprinska I, Mendes-Moreira J,
Branco P, eds. Machine Learning and Principles and Practice of Knowledge Discovery
in Databases. International Workshops of ECML PKDD 2025, Porto, Portugal, September
15–19, 2025, Revised Selected Papers, Part IV. Communications in Computer
and Information Science. Cham: Springer Nature Switzerland; 2026:5-21. doi:10.1007/978-3-032-19105-2_1'
apa: 'Jaster, B., Tharwat, A., Sheikh, E. M., Kohlhase, M., & Schenck, W. (2026).
Low Query Budget Active Learning for Classification and Regression. In I. Koprinska,
J. Mendes-Moreira, & P. Branco (Eds.), Machine Learning and Principles
and Practice of Knowledge Discovery in Databases. International Workshops of ECML
PKDD 2025, Porto, Portugal, September 15–19, 2025, Revised Selected Papers, Part
IV (pp. 5–21). Cham: Springer Nature Switzerland. https://doi.org/10.1007/978-3-032-19105-2_1'
bibtex: '@inproceedings{Jaster_Tharwat_Sheikh_Kohlhase_Schenck_2026, place={Cham},
series={Communications in Computer and Information Science}, title={Low Query
Budget Active Learning for Classification and Regression}, DOI={10.1007/978-3-032-19105-2_1},
booktitle={Machine Learning and Principles and Practice of Knowledge Discovery
in Databases. International Workshops of ECML PKDD 2025, Porto, Portugal, September
15–19, 2025, Revised Selected Papers, Part IV}, publisher={Springer Nature Switzerland},
author={Jaster, Bjarne and Tharwat, Alaa and Sheikh, Eiram Mahera and Kohlhase,
Martin and Schenck, Wolfram}, editor={Koprinska, Irena and Mendes-Moreira, João
and Branco, PaulaEditors}, year={2026}, pages={5–21}, collection={Communications
in Computer and Information Science} }'
chicago: 'Jaster, Bjarne, Alaa Tharwat, Eiram Mahera Sheikh, Martin Kohlhase, and
Wolfram Schenck. “Low Query Budget Active Learning for Classification and Regression.”
In Machine Learning and Principles and Practice of Knowledge Discovery in Databases.
International Workshops of ECML PKDD 2025, Porto, Portugal, September 15–19, 2025,
Revised Selected Papers, Part IV, edited by Irena Koprinska, João Mendes-Moreira,
and Paula Branco, 5–21. Communications in Computer and Information Science. Cham:
Springer Nature Switzerland, 2026. https://doi.org/10.1007/978-3-032-19105-2_1.'
ieee: B. Jaster, A. Tharwat, E. M. Sheikh, M. Kohlhase, and W. Schenck, “Low Query
Budget Active Learning for Classification and Regression,” in Machine Learning
and Principles and Practice of Knowledge Discovery in Databases. International
Workshops of ECML PKDD 2025, Porto, Portugal, September 15–19, 2025, Revised Selected
Papers, Part IV, Porto, Portugal, 2026, pp. 5–21.
mla: Jaster, Bjarne, et al. “Low Query Budget Active Learning for Classification
and Regression.” Machine Learning and Principles and Practice of Knowledge
Discovery in Databases. International Workshops of ECML PKDD 2025, Porto, Portugal,
September 15–19, 2025, Revised Selected Papers, Part IV, edited by Irena Koprinska
et al., Springer Nature Switzerland, 2026, pp. 5–21, doi:10.1007/978-3-032-19105-2_1.
short: 'B. Jaster, A. Tharwat, E.M. Sheikh, M. Kohlhase, W. Schenck, in: I. Koprinska,
J. Mendes-Moreira, P. Branco (Eds.), Machine Learning and Principles and Practice
of Knowledge Discovery in Databases. International Workshops of ECML PKDD 2025,
Porto, Portugal, September 15–19, 2025, Revised Selected Papers, Part IV, Springer
Nature Switzerland, Cham, 2026, pp. 5–21.'
conference:
end_date: 2025-09-19
location: Porto, Portugal
name: ECML PKDD 2025
start_date: 2025-09-15
date_created: 2026-05-10T07:55:29Z
date_updated: 2026-05-11T08:24:11Z
doi: 10.1007/978-3-032-19105-2_1
editor:
- first_name: Irena
full_name: Koprinska, Irena
last_name: Koprinska
- first_name: João
full_name: Mendes-Moreira, João
last_name: Mendes-Moreira
- first_name: Paula
full_name: Branco, Paula
last_name: Branco
language:
- iso: eng
main_file_link:
- open_access: '1'
url: https://rdcu.be/fh03T
oa: '1'
page: 5-21
place: Cham
project:
- _id: f432a2ee-bceb-11ed-a251-a83585c5074d
name: Institute for Data Science Solutions
publication: Machine Learning and Principles and Practice of Knowledge Discovery in
Databases. International Workshops of ECML PKDD 2025, Porto, Portugal, September
15–19, 2025, Revised Selected Papers, Part IV
publication_identifier:
eisbn:
- 978-3-032-19105-2
eissn:
- 1865-0937
isbn:
- 978-3-032-19104-5
issn:
- 1865-0929
publication_status: published
publisher: Springer Nature Switzerland
series_title: Communications in Computer and Information Science
status: public
title: Low Query Budget Active Learning for Classification and Regression
type: conference
user_id: '220548'
year: '2026'
...