Explaining Multi-objective Multispectral Models for Soil Parameters’ Retrieval

Jakub Mielcarek; Katarzyna Rudnik; Adriana Niepala; Mina Bikhit; Jakub Nalepa; Bogdan Ruszczak

doi:10.1007/978-3-032-36806-5_3

Explaining Multi-objective Multispectral Models for Soil Parameters’ Retrieval

Jakub Mielcarek
, Katarzyna Rudnik
, Adriana Niepala
, Mina Bikhit
, Jakub Nalepa
, Bogdan Ruszczak

Department of Algorithmics and Software (RAU/RAU5)

Opole University of Technology
Faculty of Space Technologies
AGH University of Krakow
KP Labs Spółka z ograniczoną odpowiedzialnością

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Estimating soil biophysical parameters from multispectral imagery is essential for precision agriculture, yet deploying accurate machine learning pipelines presents significant computational challenges. Individual soil trace elements often require specific model configurations, making traditional exhaustive hyperparameter optimization methods prohibitively time-consuming for resource-constrained embedded systems. In this study, we present a rigorous comparative analysis of two hyperparameter optimization strategies for multi-target soil estimation: the conventional, element-specific grid search, and the multi-objective Non-dominated Sorting Genetic Algorithm II (NSGA-II). Moreover, three explainable artificial intelligence methods were employed to analyze the behavior of the developed models and to provide insights into the contribution of input features to their predictions.

Original language	English
Title of host publication	Explainable AI in Space - 2nd International Workshop, EASi 2026, Held in Conjunction with the 35th International Joint Conference on Artificial Intelligence, IJCAI-ECAI 2026, Proceedings
Editors	Jakub Nalepa, Lukasz Tulczyjew, Agata M. Wijata, Przemyslaw Biecek, Krzysztof Kotowski, Marek Kraft, Nicolas Longépé, Evridiki V. Ntagiou
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	27-45
Number of pages	19
ISBN (Print)	9783032368058
DOIs	https://doi.org/10.1007/978-3-032-36806-5_3
Publication status	Published - 2027
Event	2nd International Workshop on Explainable AI in Space, EASi 2026, Held in Conjunction with the 35th International Joint Conference on Artificial Intelligence, IJCAI-ECAI 2026 - Bremen, Germany Duration: 15 Aug 2026 → 21 Aug 2026

Publication series

Name	Communications in Computer and Information Science
Volume	3107 CCIS
ISSN (Print)	1865-0929
ISSN (Electronic)	1865-0937

Conference

Conference	2nd International Workshop on Explainable AI in Space, EASi 2026, Held in Conjunction with the 35th International Joint Conference on Artificial Intelligence, IJCAI-ECAI 2026
Country/Territory	Germany
City	Bremen
Period	15/08/26 → 21/08/26

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 2 Zero Hunger

Keywords

explainable artificial intelligence
machine learning
non-dominated sorting genetic algorithm
precision agriculture
remote sensing
soil parameters’ estimation

ASJC Scopus subject areas

General Computer Science
General Mathematics

Access to Document

10.1007/978-3-032-36806-5_3

Cite this

Mielcarek, J., Rudnik, K., Niepala, A., Bikhit, M., Nalepa, J., & Ruszczak, B. (2027). Explaining Multi-objective Multispectral Models for Soil Parameters’ Retrieval. In J. Nalepa, L. Tulczyjew, A. M. Wijata, P. Biecek, K. Kotowski, M. Kraft, N. Longépé, & E. V. Ntagiou (Eds.), Explainable AI in Space - 2nd International Workshop, EASi 2026, Held in Conjunction with the 35th International Joint Conference on Artificial Intelligence, IJCAI-ECAI 2026, Proceedings (pp. 27-45). (Communications in Computer and Information Science; Vol. 3107 CCIS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-032-36806-5_3

Mielcarek, Jakub ; Rudnik, Katarzyna ; Niepala, Adriana et al. / Explaining Multi-objective Multispectral Models for Soil Parameters’ Retrieval. Explainable AI in Space - 2nd International Workshop, EASi 2026, Held in Conjunction with the 35th International Joint Conference on Artificial Intelligence, IJCAI-ECAI 2026, Proceedings. editor / Jakub Nalepa ; Lukasz Tulczyjew ; Agata M. Wijata ; Przemyslaw Biecek ; Krzysztof Kotowski ; Marek Kraft ; Nicolas Longépé ; Evridiki V. Ntagiou. Springer Science and Business Media Deutschland GmbH, 2027. pp. 27-45 (Communications in Computer and Information Science).

@inproceedings{c243e7ba226d4bfb8dc97c906c74a0f3,

title = "Explaining Multi-objective Multispectral Models for Soil Parameters{\textquoteright} Retrieval",

abstract = "Estimating soil biophysical parameters from multispectral imagery is essential for precision agriculture, yet deploying accurate machine learning pipelines presents significant computational challenges. Individual soil trace elements often require specific model configurations, making traditional exhaustive hyperparameter optimization methods prohibitively time-consuming for resource-constrained embedded systems. In this study, we present a rigorous comparative analysis of two hyperparameter optimization strategies for multi-target soil estimation: the conventional, element-specific grid search, and the multi-objective Non-dominated Sorting Genetic Algorithm II (NSGA-II). Moreover, three explainable artificial intelligence methods were employed to analyze the behavior of the developed models and to provide insights into the contribution of input features to their predictions.",

keywords = "explainable artificial intelligence, machine learning, non-dominated sorting genetic algorithm, precision agriculture, remote sensing, soil parameters{\textquoteright} estimation",

author = "Jakub Mielcarek and Katarzyna Rudnik and Adriana Niepala and Mina Bikhit and Jakub Nalepa and Bogdan Ruszczak",

note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Switzerland AG 2027.; 2nd International Workshop on Explainable AI in Space, EASi 2026, Held in Conjunction with the 35th International Joint Conference on Artificial Intelligence, IJCAI-ECAI 2026 ; Conference date: 15-08-2026 Through 21-08-2026",

year = "2027",

doi = "10.1007/978-3-032-36806-5\_3",

language = "English",

isbn = "9783032368058",

series = "Communications in Computer and Information Science",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "27--45",

editor = "Jakub Nalepa and Lukasz Tulczyjew and Wijata, \{Agata M.\} and Przemyslaw Biecek and Krzysztof Kotowski and Marek Kraft and Nicolas Long{\'e}p{\'e} and Ntagiou, \{Evridiki V.\}",

booktitle = "Explainable AI in Space - 2nd International Workshop, EASi 2026, Held in Conjunction with the 35th International Joint Conference on Artificial Intelligence, IJCAI-ECAI 2026, Proceedings",

address = "Germany",

}

Mielcarek, J, Rudnik, K, Niepala, A, Bikhit, M, Nalepa, J & Ruszczak, B 2027, Explaining Multi-objective Multispectral Models for Soil Parameters’ Retrieval. in J Nalepa, L Tulczyjew, AM Wijata, P Biecek, K Kotowski, M Kraft, N Longépé & EV Ntagiou (eds), Explainable AI in Space - 2nd International Workshop, EASi 2026, Held in Conjunction with the 35th International Joint Conference on Artificial Intelligence, IJCAI-ECAI 2026, Proceedings. Communications in Computer and Information Science, vol. 3107 CCIS, Springer Science and Business Media Deutschland GmbH, pp. 27-45, 2nd International Workshop on Explainable AI in Space, EASi 2026, Held in Conjunction with the 35th International Joint Conference on Artificial Intelligence, IJCAI-ECAI 2026, Bremen, Germany, 15/08/26. https://doi.org/10.1007/978-3-032-36806-5_3

Explaining Multi-objective Multispectral Models for Soil Parameters’ Retrieval. / Mielcarek, Jakub; Rudnik, Katarzyna; Niepala, Adriana et al.
Explainable AI in Space - 2nd International Workshop, EASi 2026, Held in Conjunction with the 35th International Joint Conference on Artificial Intelligence, IJCAI-ECAI 2026, Proceedings. ed. / Jakub Nalepa; Lukasz Tulczyjew; Agata M. Wijata; Przemyslaw Biecek; Krzysztof Kotowski; Marek Kraft; Nicolas Longépé; Evridiki V. Ntagiou. Springer Science and Business Media Deutschland GmbH, 2027. p. 27-45 (Communications in Computer and Information Science; Vol. 3107 CCIS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Explaining Multi-objective Multispectral Models for Soil Parameters’ Retrieval

AU - Mielcarek, Jakub

AU - Rudnik, Katarzyna

AU - Niepala, Adriana

AU - Bikhit, Mina

AU - Nalepa, Jakub

AU - Ruszczak, Bogdan

N1 - Publisher Copyright: © The Author(s), under exclusive license to Springer Nature Switzerland AG 2027.

PY - 2027

Y1 - 2027

N2 - Estimating soil biophysical parameters from multispectral imagery is essential for precision agriculture, yet deploying accurate machine learning pipelines presents significant computational challenges. Individual soil trace elements often require specific model configurations, making traditional exhaustive hyperparameter optimization methods prohibitively time-consuming for resource-constrained embedded systems. In this study, we present a rigorous comparative analysis of two hyperparameter optimization strategies for multi-target soil estimation: the conventional, element-specific grid search, and the multi-objective Non-dominated Sorting Genetic Algorithm II (NSGA-II). Moreover, three explainable artificial intelligence methods were employed to analyze the behavior of the developed models and to provide insights into the contribution of input features to their predictions.

AB - Estimating soil biophysical parameters from multispectral imagery is essential for precision agriculture, yet deploying accurate machine learning pipelines presents significant computational challenges. Individual soil trace elements often require specific model configurations, making traditional exhaustive hyperparameter optimization methods prohibitively time-consuming for resource-constrained embedded systems. In this study, we present a rigorous comparative analysis of two hyperparameter optimization strategies for multi-target soil estimation: the conventional, element-specific grid search, and the multi-objective Non-dominated Sorting Genetic Algorithm II (NSGA-II). Moreover, three explainable artificial intelligence methods were employed to analyze the behavior of the developed models and to provide insights into the contribution of input features to their predictions.

KW - explainable artificial intelligence

KW - machine learning

KW - non-dominated sorting genetic algorithm

KW - precision agriculture

KW - remote sensing

KW - soil parameters’ estimation

UR - https://www.scopus.com/pages/publications/105048686860

U2 - 10.1007/978-3-032-36806-5_3

DO - 10.1007/978-3-032-36806-5_3

M3 - Conference contribution

AN - SCOPUS:105048686860

SN - 9783032368058

T3 - Communications in Computer and Information Science

SP - 27

EP - 45

BT - Explainable AI in Space - 2nd International Workshop, EASi 2026, Held in Conjunction with the 35th International Joint Conference on Artificial Intelligence, IJCAI-ECAI 2026, Proceedings

A2 - Nalepa, Jakub

A2 - Tulczyjew, Lukasz

A2 - Wijata, Agata M.

A2 - Biecek, Przemyslaw

A2 - Kotowski, Krzysztof

A2 - Kraft, Marek

A2 - Longépé, Nicolas

A2 - Ntagiou, Evridiki V.

PB - Springer Science and Business Media Deutschland GmbH

T2 - 2nd International Workshop on Explainable AI in Space, EASi 2026, Held in Conjunction with the 35th International Joint Conference on Artificial Intelligence, IJCAI-ECAI 2026

Y2 - 15 August 2026 through 21 August 2026

ER -

Mielcarek J, Rudnik K, Niepala A, Bikhit M, Nalepa J, Ruszczak B. Explaining Multi-objective Multispectral Models for Soil Parameters’ Retrieval. In Nalepa J, Tulczyjew L, Wijata AM, Biecek P, Kotowski K, Kraft M, Longépé N, Ntagiou EV, editors, Explainable AI in Space - 2nd International Workshop, EASi 2026, Held in Conjunction with the 35th International Joint Conference on Artificial Intelligence, IJCAI-ECAI 2026, Proceedings. Springer Science and Business Media Deutschland GmbH. 2027. p. 27-45. (Communications in Computer and Information Science). doi: 10.1007/978-3-032-36806-5_3

Explaining Multi-objective Multispectral Models for Soil Parameters’ Retrieval

Abstract

Publication series

Conference

UN SDGs

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this