Statistically Aware Optimization for Resource-constrained and Geometrically-rich Data: Nigerian Agricultural Case Study

Ogethakpo Arhonefe Joseph; Ovbije Oghenekevwe Godspower; Adelakun Abdul Olawale; Ejakpovi Simeon Uyovwieyovwe; Emunefe Friday Gabriel; Akworigbe Hope Avwerosuoghene; Akporavware Oforhoraye Sunday

doi:doi:10.11648/j.ijssam.20251003.12

Research Article |

| Peer-Reviewed

Statistically Aware Optimization for Resource-constrained and Geometrically-rich Data: Nigerian Agricultural Case Study

Ogethakpo Arhonefe Joseph^*

, Ovbije Oghenekevwe Godspower

, Adelakun Abdul Olawale

, Ejakpovi Simeon Uyovwieyovwe

, Emunefe Friday Gabriel

, Akworigbe Hope Avwerosuoghene

, Akporavware Oforhoraye Sunday

Published in International Journal of Systems Science and Applied Mathematics (Volume 10, Issue 3)

Received: 16 September 2025 Accepted: 26 September 2025 Published: 26 November 2025

Views: Downloads:

Download PDF

Share This Article

Twitter
Linked In
Facebook

Abstract

Modern machine learning, fueled by large datasets and complex models, faces a critical tension. The statistical principles underpinning learning (generalization, efficiency, robustness) often clash with the computational realities of optimization, especially in a resource constrained environment or when data exhibits inherent geometric structure. This work addresses the theme "Statistics Meets Optimization" by employing an optimization framework explicitly designed to leverage statistical data properties, particularly group invariances/equivariances common in real world data (e.g., spatial rotations in satellite imagery, temporal shifts in sensor data), to achieve significant gains in sample efficiency and convergence speed. We theoretically derive generalization bounds linking the exploitation of data geometry to reduced sample complexity. Empirically, we demonstrate the efficacy of our method on a challenging real world case study, i.e., on predicting crop yield anomalies in Delta State, Nigeria, using limited, noisy, and spatially heterogeneous satellite and meteorological data. Our optimizer achieved a significant performance with 40% less data compared to adaptive baselines (Adam, RMSProp), highlighting the practical impact of statistically-informed optimization, especially for regions facing data scarcity. This work provides a concrete bridge between statistical theory (data structure, efficiency) and optimization practice (algorithm design, scalability), demonstrating that geometry-aware algorithms can democratize effective ML for resource-limited applications.

Published in	International Journal of Systems Science and Applied Mathematics (Volume 10, Issue 3)
DOI	10.11648/j.ijssam.20251003.12
Page(s)	46-57
Creative Commons	This is an Open Access article, distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution and reproduction in any medium or format, provided the original work is properly cited.
Copyright	Copyright © The Author(s), 2025. Published by Science Publishing Group

Keywords

Geometry-Aware Optimization, Statistical Learning, GeoAda, Crop Yield Prediction, Data Scarcity, Resource-constrained Machine Learning, Nigeria

References

[1]	Ruder, S. (2016). An overview of gradient descent optimization algorithms. arXiv: 1609.04747 https://doi.org/10.48550/arXiv.1609.04747
[2]	GeeksforGeeks. (2025, July 23). Optimization algorithms in machine Learning. https://www.geeksforgeeks.org/machine-learning/optimization-algorithms-in-machine-learning/
[3]	Ogethakpo, A. J., Ovbije, O. G., and Apanapudor, J. S. (2025). The Evolving Landscape of Optimization: Current Trends and Future Directions. International Journal of Modern Science and Research Technology. 3(8), 108-115. https://doi.org/10.5281/zenodo.16894145
[4]	Karthick, K. (2024). Comprehensive Overview of Optimization Techniques in Machine Learning Training. Control Systems and Optimization Letters, 2(1) https://doi.org/10.59247/csol.v2i1.69
[5]	Goodfellow, I., Bengio, Y., and Courville, A. (2016). Deep Learning. MIT Press.
[6]	Bartlett, P., Foster, D. J., and Telgarsky, M. (2017). Spectrally-normalized margin bounds for neural networks. Advances in Neural Information Processing System. vol. 30. https://doi.org/10.48550/arXiv.1706.08498
[7]	Sabrepc Blog. (July 10, 2025). Overparameterization in AI Models - More Parameters Never Hurts. https://www.sabrepc.com/blog/deep-learning-and-ai/overparameterization-in-ai-models
[8]	Du, S., Lee, J., Li, H., Wang, L., and Zhai, X. (2019). Gradient descent finds global minima of deep neural networks. International Conference on Machine Learning (ICML), PMLR 97: 1675-1685.
[9]	Maleki, F., Ovens, K., Gupta, R., Reinhold, C., Spatz, A., & Forghani, R. (2022). Generalizability of Machine Learning Models: Quantitative Evaluation of Three Methodological Pitfalls. Radiology. Artificial intelligence, 5(1), e220028. https://doi.org/10.1148/ryai.220028
[10]	Stephens (2023). Challenges and Solutions in Training Deep Learning Models. Data Science Society, Data. Platform. https://www.datasciencesociety.net/challenges-and-solutions-in-training-deep-learning-models/
[11]	Hu, N. (2022). An Empirical Study of Neural Network Training Dynamics. Medium. https://medium.com/@hu.niel92/an-empirical-study-of-neural-network-training-dynamics-3bf4b659359e
[12]	Song, P. (2024). Machine Learning Scalability Issues: Challenges and Solutions. ML Journey. https://mljourney.com/machine-learning-scalability-issues-challenges-and-solutions/
[13]	Belkin, M., Hsu, D., Ma, S., and Mandal, S. (2019). Reconciling Modern Machine Learning Practice and the Classical Bias-Variance Trade-off. Proceedings of the National Academy of Science, U.S.A. 116(32) 15849-15854 https://doi.org/10.1073/pnas.1903070116
[14]	Neyshabur, B., Bhojanapalli, S., McAllester, D., and Srebro, N. (2017). Exploring generalization in deep learning. NIPS’17: Proceedings of the 31^st International Conference on Neural Information Processing System. 30, 5949-5958 https://doi.org/10.48550/arXiv.1706.08947
[15]	Pan, S. J. and Yang, Q. (2010). A survey on transfer learning. IEEE Transactions on Knowledge and Data Engineering. 22(10), 1345–1359. https://doi.org/10.1109/TKDE.2009.191
[16]	Nair, V. and Hinton, G. E. (2010). Rectified linear units improve restricted Boltzmann machines. ICML’10: Proceedings of the 27^th International Conference on Machine Learning., pp. 807-814.
[17]	Ogethakpo, A. J., & Ojobor, S. A. (2021). The Lotka-Volterra Predator-Prey Model with Disturbance. Nigerian Journal of Science and Environment. 19(2), 135-144.
[18]	Ogethakpo, A. J., Ogoegbulem, O., Apanapudor, J. S., & Sanubi, H. O. (2025). Stochastic Dynamics of Urban Predator-Prey Systems: Integrating Human Disturbance and Functional Responses. American Journal of Applied Mathematics. 13(4), 282-291. https://doi.org/10.11648/j.ajam.20251304.15
[19]	Kingma, D. and Ba, J. (2015). Adam: A Method for Stochastic Optimization. Proceedings of the 3rd International Conference on Learning Representations (ICLR 2015). https://doi.org/10.48550/arXiv.1412.6980
[20]	T. Tieleman and G. Hinton. Lecture 6.5-RMSProp: Divide the Gradient by a Running Average of its Recent Magnitude. COURSERA: Neural Networks for Machine Learning. 4(2), 26-31.
[21]	Symmetry. (2021, December 14). Department of Mathematics at UTSA,. Retrieved 09: 35, September 11, 2025 from https://mathresearch.utsa.edu/wiki/index.php?title=Symmetry&oldid=4194
[22]	Shao, H., Montasser, O., and Blum, A. (2022). A Theory of PAC Learnability under Transformation Invariances. Cornell University. arXiv. https://arxiv.org/abs/2202.07552
[23]	Cohen, T. and Welling, M. (2016). Group Equivariant Convolutional Networks. International Conference on Machine Learning (ICML). 2990-2999. https://doi.org/10.48550/arXiv.1602.07576
[24]	Maurer, A. (2006). The Rademacher complexity of linear transformation classes. International Conference on Computational Learning Theory. pp. 65-78. https://doi.org/10.1007/11776420_8
[25]	Ovbije, O. G., Oyiborhoro, M., and Akworigbe, A. H. Optimizing Agricultural Stock Portfolios in Ughelli Town Using Linear Programming. Faculty of Natural and Applied Sciences Journal of Mathematical Modeling and Numerical Simulation. 2(1), 108-114.
[26]	Lobell, D. B. (2013). The use of satellite data for crop yield gap analysis. Field Crops Research. 143, 56-64. https://doi.org/10.1016/j.fcr.2012.08.008
[27]	Krizhevsky, A., Sutskever, I. and Hinton, G. E. (2017). ImageNet classification with deep convolutional neural networks. Communications of the ACM. 60(6), 84-90. https://doi.org/10.1145/3065386
[28]	Bronstein, M., Bruna, J., Cohen, T., and Velivckovi, P. (2021). Geometric deep learning: Grids, groups, graphs, geodesics, and gauges. arXiv, abs/2104.13478.
[29]	Vapnik, V. N. (1995). The Nature of Statistical Learning Theory. Springer, New York. http://dx.doi.org/10.1007/978-1-4757-2440-0
[30]	LeCun, Y., Bengio, Y., and Hinton, G. (2015). Deep learning. Nature. 521(7553), 436-444. https://doi.org/10.1038/nature14539
[31]	Kawaguchi, K., Bengio, Y., and Kaelbling, L. (2022). Generalization in Deep Learning. Mathematical Aspects of Deep Learning. Cambridge University Press.112-148. https://doi.org/10.1017/9781009025096.003
[32]	Shalev-Shwartzc, S. and Ben-David, S. (2014). Understanding Machine Learning: From Theory to Algorithms. Cambridge Univ. Press. https://doi.org/10.1017/CBO9781107298019
[33]	Keskar, N. S., Mudigere, D., Nocedal, J., Smelyanskiy, M., and Tang, P. T. P. (2017). On large-batch training for deep learning: Generalization gap and sharp minima. International Conference on Learning Representations. https://doi.org/10.48550/arXiv.1609.04836
[34]	Robbins, H. and Monro, S. (1951). A stochastic approximation method. The Annals of Mathematical Statistics. 22(3), 400-407. https://www.jstor.org/stable/2236626
[35]	Bousquet, O. and Elisseeff, A. (2002). Stability and Generalization. Journal of Machine Learning Research. 2, 499-526.
[36]	Hardt, M., Recht, B., and Singer, Y. (2016). Train faster, generalize better: Stability of stochastic gradient descent. International Conference on Machine Learning. JMLR: W&CP 48, 1225-1234. https://doi.org/10.48550/arXiv.1509.01240
[37]	Roberts, D. A., Yaida, S. and Hanin, B. (2022). The Principles of Deep Learning Theory: An Effective Theory Approach to Understanding Neural Networks. Cambridge Univ. Press. https://doi.org/10.1017/9781009023405

Cite This Article

Plain Text BibTeX RIS

APA Style

Joseph, O. A., Godspower, O. O., Olawale, A. A., Uyovwieyovwe, E. S., Gabriel, E. F., et al. (2025). Statistically Aware Optimization for Resource-constrained and Geometrically-rich Data: Nigerian Agricultural Case Study. International Journal of Systems Science and Applied Mathematics, 10(3), 46-57. https://doi.org/10.11648/j.ijssam.20251003.12

Copy | Download

ACS Style

Joseph, O. A.; Godspower, O. O.; Olawale, A. A.; Uyovwieyovwe, E. S.; Gabriel, E. F., et al. Statistically Aware Optimization for Resource-constrained and Geometrically-rich Data: Nigerian Agricultural Case Study. Int. J. Syst. Sci. Appl. Math. 2025, 10(3), 46-57. doi: 10.11648/j.ijssam.20251003.12

Copy | Download

AMA Style

Joseph OA, Godspower OO, Olawale AA, Uyovwieyovwe ES, Gabriel EF, et al. Statistically Aware Optimization for Resource-constrained and Geometrically-rich Data: Nigerian Agricultural Case Study. Int J Syst Sci Appl Math. 2025;10(3):46-57. doi: 10.11648/j.ijssam.20251003.12

Copy | Download

@article{10.11648/j.ijssam.20251003.12,
  author = {Ogethakpo Arhonefe Joseph and Ovbije Oghenekevwe Godspower and Adelakun Abdul Olawale and Ejakpovi Simeon Uyovwieyovwe and Emunefe Friday Gabriel and Akworigbe Hope Avwerosuoghene and Akporavware Oforhoraye Sunday},
  title = {Statistically Aware Optimization for Resource-constrained and Geometrically-rich Data: Nigerian Agricultural Case Study
},
  journal = {International Journal of Systems Science and Applied Mathematics},
  volume = {10},
  number = {3},
  pages = {46-57},
  doi = {10.11648/j.ijssam.20251003.12},
  url = {https://doi.org/10.11648/j.ijssam.20251003.12},
  eprint = {https://article.sciencepublishinggroup.com/pdf/10.11648.j.ijssam.20251003.12},
  abstract = {Modern machine learning, fueled by large datasets and complex models, faces a critical tension. The statistical principles underpinning learning (generalization, efficiency, robustness) often clash with the computational realities of optimization, especially in a resource constrained environment or when data exhibits inherent geometric structure. This work addresses the theme "Statistics Meets Optimization" by employing an optimization framework explicitly designed to leverage statistical data properties, particularly group invariances/equivariances common in real world data (e.g., spatial rotations in satellite imagery, temporal shifts in sensor data), to achieve significant gains in sample efficiency and convergence speed. We theoretically derive generalization bounds linking the exploitation of data geometry to reduced sample complexity. Empirically, we demonstrate the efficacy of our method on a challenging real world case study, i.e., on predicting crop yield anomalies in Delta State, Nigeria, using limited, noisy, and spatially heterogeneous satellite and meteorological data. Our optimizer achieved a significant performance with 40% less data compared to adaptive baselines (Adam, RMSProp), highlighting the practical impact of statistically-informed optimization, especially for regions facing data scarcity. This work provides a concrete bridge between statistical theory (data structure, efficiency) and optimization practice (algorithm design, scalability), demonstrating that geometry-aware algorithms can democratize effective ML for resource-limited applications.},
 year = {2025}
}

Copy | Download

TY  - JOUR
T1  - Statistically Aware Optimization for Resource-constrained and Geometrically-rich Data: Nigerian Agricultural Case Study

AU  - Ogethakpo Arhonefe Joseph
AU  - Ovbije Oghenekevwe Godspower
AU  - Adelakun Abdul Olawale
AU  - Ejakpovi Simeon Uyovwieyovwe
AU  - Emunefe Friday Gabriel
AU  - Akworigbe Hope Avwerosuoghene
AU  - Akporavware Oforhoraye Sunday
Y1  - 2025/11/26
PY  - 2025
N1  - https://doi.org/10.11648/j.ijssam.20251003.12
DO  - 10.11648/j.ijssam.20251003.12
T2  - International Journal of Systems Science and Applied Mathematics
JF  - International Journal of Systems Science and Applied Mathematics
JO  - International Journal of Systems Science and Applied Mathematics
SP  - 46
EP  - 57
PB  - Science Publishing Group
SN  - 2575-5803
UR  - https://doi.org/10.11648/j.ijssam.20251003.12
AB  - Modern machine learning, fueled by large datasets and complex models, faces a critical tension. The statistical principles underpinning learning (generalization, efficiency, robustness) often clash with the computational realities of optimization, especially in a resource constrained environment or when data exhibits inherent geometric structure. This work addresses the theme "Statistics Meets Optimization" by employing an optimization framework explicitly designed to leverage statistical data properties, particularly group invariances/equivariances common in real world data (e.g., spatial rotations in satellite imagery, temporal shifts in sensor data), to achieve significant gains in sample efficiency and convergence speed. We theoretically derive generalization bounds linking the exploitation of data geometry to reduced sample complexity. Empirically, we demonstrate the efficacy of our method on a challenging real world case study, i.e., on predicting crop yield anomalies in Delta State, Nigeria, using limited, noisy, and spatially heterogeneous satellite and meteorological data. Our optimizer achieved a significant performance with 40% less data compared to adaptive baselines (Adam, RMSProp), highlighting the practical impact of statistically-informed optimization, especially for regions facing data scarcity. This work provides a concrete bridge between statistical theory (data structure, efficiency) and optimization practice (algorithm design, scalability), demonstrating that geometry-aware algorithms can democratize effective ML for resource-limited applications.
VL  - 10
IS  - 3
ER  -

Copy | Download

Author Information

Ogethakpo Arhonefe Joseph

Department of Mathematics, Delta State University, Abraka, Nigeria

Contact Email

http://orcid.org/0009-0004-6013-1341
Ovbije Oghenekevwe Godspower

Department of Mathematics, Delta State University, Abraka, Nigeria

http://orcid.org/0009-0004-3928-1740
Adelakun Abdul Olawale

Department of Computer and Industrial & Production Engineering, Abiola Ajimobi Technical University, Ibadan, Nigeria

http://orcid.org/0009-0006-7468-1744
Ejakpovi Simeon Uyovwieyovwe

Department of Mathematics, College of Education, Warri, Nigeria

http://orcid.org/0009-0008-7425-4966
Emunefe Friday Gabriel

Department of Mathematics, College of Education, Mosogar, Nigeria

http://orcid.org/0009-0005-1892-8303
Akworigbe Hope Avwerosuoghene

Department of Mathematics, Federal University of Petroleum Resources, Effurun, Nigeria

http://orcid.org/0009-0006-0902-7455
Akporavware Oforhoraye Sunday

Department of Mathematics, College of Education, Mosogar, Nigeria

http://orcid.org/0009-0000-0366-6076

Download PDF

Submit an Article

Sections

Plain Text BibTeX RIS

APA Style

Joseph, O. A., Godspower, O. O., Olawale, A. A., Uyovwieyovwe, E. S., Gabriel, E. F., et al. (2025). Statistically Aware Optimization for Resource-constrained and Geometrically-rich Data: Nigerian Agricultural Case Study. International Journal of Systems Science and Applied Mathematics, 10(3), 46-57. https://doi.org/10.11648/j.ijssam.20251003.12

Copy | Download

ACS Style

Joseph, O. A.; Godspower, O. O.; Olawale, A. A.; Uyovwieyovwe, E. S.; Gabriel, E. F., et al. Statistically Aware Optimization for Resource-constrained and Geometrically-rich Data: Nigerian Agricultural Case Study. Int. J. Syst. Sci. Appl. Math. 2025, 10(3), 46-57. doi: 10.11648/j.ijssam.20251003.12

Copy | Download

AMA Style

Joseph OA, Godspower OO, Olawale AA, Uyovwieyovwe ES, Gabriel EF, et al. Statistically Aware Optimization for Resource-constrained and Geometrically-rich Data: Nigerian Agricultural Case Study. Int J Syst Sci Appl Math. 2025;10(3):46-57. doi: 10.11648/j.ijssam.20251003.12

Copy | Download

@article{10.11648/j.ijssam.20251003.12,
  author = {Ogethakpo Arhonefe Joseph and Ovbije Oghenekevwe Godspower and Adelakun Abdul Olawale and Ejakpovi Simeon Uyovwieyovwe and Emunefe Friday Gabriel and Akworigbe Hope Avwerosuoghene and Akporavware Oforhoraye Sunday},
  title = {Statistically Aware Optimization for Resource-constrained and Geometrically-rich Data: Nigerian Agricultural Case Study
},
  journal = {International Journal of Systems Science and Applied Mathematics},
  volume = {10},
  number = {3},
  pages = {46-57},
  doi = {10.11648/j.ijssam.20251003.12},
  url = {https://doi.org/10.11648/j.ijssam.20251003.12},
  eprint = {https://article.sciencepublishinggroup.com/pdf/10.11648.j.ijssam.20251003.12},
  abstract = {Modern machine learning, fueled by large datasets and complex models, faces a critical tension. The statistical principles underpinning learning (generalization, efficiency, robustness) often clash with the computational realities of optimization, especially in a resource constrained environment or when data exhibits inherent geometric structure. This work addresses the theme "Statistics Meets Optimization" by employing an optimization framework explicitly designed to leverage statistical data properties, particularly group invariances/equivariances common in real world data (e.g., spatial rotations in satellite imagery, temporal shifts in sensor data), to achieve significant gains in sample efficiency and convergence speed. We theoretically derive generalization bounds linking the exploitation of data geometry to reduced sample complexity. Empirically, we demonstrate the efficacy of our method on a challenging real world case study, i.e., on predicting crop yield anomalies in Delta State, Nigeria, using limited, noisy, and spatially heterogeneous satellite and meteorological data. Our optimizer achieved a significant performance with 40% less data compared to adaptive baselines (Adam, RMSProp), highlighting the practical impact of statistically-informed optimization, especially for regions facing data scarcity. This work provides a concrete bridge between statistical theory (data structure, efficiency) and optimization practice (algorithm design, scalability), demonstrating that geometry-aware algorithms can democratize effective ML for resource-limited applications.},
 year = {2025}
}

Copy | Download

TY  - JOUR
T1  - Statistically Aware Optimization for Resource-constrained and Geometrically-rich Data: Nigerian Agricultural Case Study

AU  - Ogethakpo Arhonefe Joseph
AU  - Ovbije Oghenekevwe Godspower
AU  - Adelakun Abdul Olawale
AU  - Ejakpovi Simeon Uyovwieyovwe
AU  - Emunefe Friday Gabriel
AU  - Akworigbe Hope Avwerosuoghene
AU  - Akporavware Oforhoraye Sunday
Y1  - 2025/11/26
PY  - 2025
N1  - https://doi.org/10.11648/j.ijssam.20251003.12
DO  - 10.11648/j.ijssam.20251003.12
T2  - International Journal of Systems Science and Applied Mathematics
JF  - International Journal of Systems Science and Applied Mathematics
JO  - International Journal of Systems Science and Applied Mathematics
SP  - 46
EP  - 57
PB  - Science Publishing Group
SN  - 2575-5803
UR  - https://doi.org/10.11648/j.ijssam.20251003.12
AB  - Modern machine learning, fueled by large datasets and complex models, faces a critical tension. The statistical principles underpinning learning (generalization, efficiency, robustness) often clash with the computational realities of optimization, especially in a resource constrained environment or when data exhibits inherent geometric structure. This work addresses the theme "Statistics Meets Optimization" by employing an optimization framework explicitly designed to leverage statistical data properties, particularly group invariances/equivariances common in real world data (e.g., spatial rotations in satellite imagery, temporal shifts in sensor data), to achieve significant gains in sample efficiency and convergence speed. We theoretically derive generalization bounds linking the exploitation of data geometry to reduced sample complexity. Empirically, we demonstrate the efficacy of our method on a challenging real world case study, i.e., on predicting crop yield anomalies in Delta State, Nigeria, using limited, noisy, and spatially heterogeneous satellite and meteorological data. Our optimizer achieved a significant performance with 40% less data compared to adaptive baselines (Adam, RMSProp), highlighting the practical impact of statistically-informed optimization, especially for regions facing data scarcity. This work provides a concrete bridge between statistical theory (data structure, efficiency) and optimization practice (algorithm design, scalability), demonstrating that geometry-aware algorithms can democratize effective ML for resource-limited applications.
VL  - 10
IS  - 3
ER  -

Copy | Download