International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 12 Issue: 10 | Oct 2025
www.irjet.net
p-ISSN: 2395-0072
A Data-Driven Approach to Agricultural Sustainability: Crop Recommendation using Machine Learning Dr.G.Arutjothi1, Dr.K.Geetha2 1 Assistant Professor, Dept. of Computer Applications, Sona College of Arts and Science,Tamil Nadu, India
2Lecturer, Dept. of Computer Applications, Government Arts College (Autonomous),Tamil Nadu, India ---------------------------------------------------------------------***--------------------------------------------------------------------Several machine learning models, such as Logistic Abstract - This paper aims to develop a crop
Regression, Support Vector Classifier (SVC), K-Neighbors Classifier, Decision Tree Classifier, Extra Tree Classifier, Random Forest Classifier, Bagging Classifier, and Gradient Boosting Classifier, were trained and assessed based on performance metrics. Among these, the Random Forest Classifier stood out for its accuracy and robustness, making it the ideal choice for implementation.
recommendation system using machine learning techniques with Python, designed to assist farmers in choosing the best crops for their fields. The system analyzes essential soil and environmental parameters such as Nitrogen (N), phosphorus (P), and Potassium (K) content in the soil, temperature, humidity, soil pH level, and rainfall. These parameters are utilized to train various machine learning models, including Logistic Regression, Support Vector Machine(SVM), KNearest Neighbor, Decision Tree, Random Forest, Bagging, and Gradient Boosting classifiers. The Random Forest Classifier was identified as the most accurate model through extensive evaluation. The system has been implemented as a user-friendly website to support agricultural officers in providing farmers with real-time data analysis for informed decision-making. This system strives to enhance farming efficiency, optimize resource usage, and improve crop yields by offering tailored crop recommendations. This paper represents a significant advancement in smart agriculture, leveraging technology to support sustainable farming practices and achieve better outcomes.
To enhance accessibility, the system is presented as an intuitive web application developed using Visual Studio. The platform allows agricultural professionals and farmers to input real-time data and receive immediate crop recommendations. Features such as user-friendly data input forms, interactive visualizations, and detailed crop analysis reports ensure that users can easily navigate and utilize the system without requiring technical expertise. The Crop Recommendation System project offers recommendations for various crops, catering to different regions and climatic conditions. Utilizing machine learning and a user-friendly website, it provides data-driven insights for farmers, enabling informed decision-making and improved crop yields. This project demonstrates the potential of technology in transforming agriculture, supporting sustainable farming practices, and contributing to food security and economic stability.This study represents a crucial advancement in promoting sustainable agriculture, enabling well-informed decision-making and contributing to enhanced crop yields and environmental conservation within the farming industry.
Keywords: Machine Learning, Recommendation System, Random Forest Classifier, Accuracy, Environmental Factors.
1.INTRODUCTION The Crop Recommendation System is an innovative initiative aimed at revolutionizing traditional farming methods by integrating advanced machine learning technologies. Its primary objective is to assist farmers in making informed, data-driven decisions regarding crop selection, which is essential for maximizing agricultural productivity and promoting sustainable practices. By utilizing machine learning, the system evaluates intricate datasets that include key soil and environmental factors, providing customized crop recommendations tailored to specific farm conditions. These parameters include soil nutrients such as Nitrogen (N), phosphorus (P), and Potassium (K), along with temperature, humidity, soil pH levels, and rainfall. The project employs Python, a highly adaptable programming language, to build and execute the machine-learning algorithms necessary for crop recommendation. Leveraging Python's powerful libraries, including scikit-learn, pandas, and numpy, the system efficiently handles data analysis and model development.
© 2025, IRJET
|
Impact Factor value: 8.315
2. LITERATURE SURVEY Grasslands, the world’s largest terrestrial ecosystem, are a vital feed source for livestock. Addressing the growing global demand for meat and dairy products sustainably poses significant challenges. Technological advancements such as the Global Positioning System (GPS) and ground-based sensors show promise for grassland and herd management. Additionally, the increasing availability of spaceborne remote sensing data highlights the need to refine methods for exploiting such imagery. Biophysical parameter retrieval for grasslands has progressed from classical regression analysis to more sophisticated modeling techniques, though high-quality calibration and validation data remain critical. The development of hyperspectral
|
ISO 9001:2008 Certified Journal
| Page 687