Loading article...
Loading article...
Stream/Area:
-
Subject:
-
Paper ID:
IJSREM66628
Volume:
10 |
8 |
Aug 2026
Pages:
1-5
Publication Frequency:
Monthly (12 issues per year)
License:International Journal of Scientific Research in Engineering and Management follows the Creative Commons Attribution 4.0 (CC BY 4.0) License, allowing free use and distribution with proper attribution.
School of Computer Application
GM University
Faculty of Computing and IT
GM University
Faculty of Computing and IT
GM University
This paper presents an AI-based soil health analytics and crop recommendation system that transforms manually entered or dataset-derived soil and climatic attributes into interpretable crop guidance. The system is intentionally designed without Internet of Things devices or hardware sensors: users provide nitrogen, phosphorus, potassium, temperature, humidity, pH, and rainfall values through a Flask web interface. A curated crop-recommendation dataset containing 2,200 records and 22 crop classes is processed using Pandas and NumPy, then evaluated using Decision Tree, Random Forest, Support Vector Machine (SVM), and K-Nearest Neighbors (KNN) classifiers. The production implementation persists a Random Forest model selected through weighted F1-score, while the other algorithms form a comparative benchmark. With a stratified 80:20 split and random seed 42, the deployed Random Forest achieved 99.32% accuracy, 99.35% weighted precision, 99.32% weighted recall, and 99.32% weighted F1-score on 440 held-out records; the Decision Tree achieved 97.95% accuracy. The workflow includes validation, preprocessing, prediction, confidence-ranked recommendations, soil-health interpretation, and report generation. Feature-importance analysis ranks humidity, rainfall, potassium, and phosphorus as influential variables, which supports transparent agronomic discussion rather than opaque label prediction. The results indicate that ensemble learning can provide accurate decision support when inputs are representative of local conditions. The paper also identifies limitations related to dataset scope, laboratory variability, and the need for region-specific validation before agronomic deployment.
KEYWORDS
This paper presented a software-only AI-based soil health analytics and crop recommendation system using manual or dataset-derived features. A Flask application operationalizes preprocessing, classification, explanation, and reporting. On the supplied 2,200-record dataset, the saved Random Forest model achieved 99.32% accuracy on a 440-record held-out set and outperformed the saved Decision Tree baseline. The solution is promising for accessible crop decision support, provided that recommendations are validated locally and used alongside agronomic expertise. The explicit separation of dataset performance from field performance is essential to responsible adoption.
[1] S. Pandian, “Crop Recommendation Dataset,” Kaggle, 2020. [Online]. Available: https://www.kaggle.com/datasets/atharvaingle/crop-recommendation-dataset [2] L. Breiman, “Random forests,” Machine Learning, vol. 45, no. 1, pp. 5–32, 2001. [3] C. Cortes and V. Vapnik, “Support-vector networks,” Machine Learning, vol. 20, no. 3, pp. 273–297, 1995. [4] T. Cover and P. Hart, “Nearest neighbor pattern classification,” IEEE Trans. Inf. Theory, vol. 13, no. 1, pp. 21–27, 1967. [5] J. R. Quinlan, C4.5: Programs for Machine Learning. San Francisco, CA, USA: Morgan Kaufmann, 1993. [6] A. Kamilaris and F. X. Prenafeta-Boldú, “Deep learning in agriculture: A survey,” Computers and Electronics in Agriculture, vol. 147, pp. 70–90, 2018. [7] M. Shahhosseini, G. Hu, I. Huber, and S. V. Archontoulis, “Coupling machine learning and crop modeling improves crop yield prediction in the US Corn Belt,” Scientific Reports, vol. 11, art. 1606, 2021. [8] F. Pedregosa et al., “Scikit-learn: Machine learning in Python,” J. Mach. Learn. Res., vol. 12, pp. 2825–2830, 2011. [9] W. McKinney, Python for Data Analysis, 2nd ed. Sebastopol, CA, USA: O’Reilly Media, 2017. [10] P. T. DeVries, J. Verbesselt, L. M. Kooistra, and M. Herold, “Robust monitoring of small-scale crop growth using UAV-derived data,” Remote Sensing, vol. 7, no. 4, pp. 5295–5316, 2015.
Publication Date
8/10/2026
Peer Review Type
Double-Blind Peer Review
Review Rounds: 2
Documents
Article Views
0
Total views since publication
PDF Downloads
2
Total downloads since publication
Geographic Reach
5
Digital Preservation: This article is permanently archived on Zenodo and complies with all data preservation standards across major indexing platforms.
Plagiarism Screening
This manuscript has been checked for plagiarism using TurnitinPlagiarismChecker software.
Conflict of Interest
The authors declare no conflict of interest related to this research work.
Funding Information
No funding information available for this research.
Related research in the same subject area
Loading similar articles...