Page 33 - TECH MAGAZINE CSE

AN
VESHAN

Crop Recommendation System using Machine Learning and IoT

Mr. Akshay Mool

Traditional methods of crop selection often rely on experience and intuition, leading to
sub-optimal yields and resource utilization. This project introduces a novel crop
recommendation system that leverages the power of machine learning and sensor
technology to revolutionize agricultural practices.

The proposed system integrates six machine learning algorithms-logistic regression,
decision tree, support vector machine (SVM), naive Bayes, random forest, and XGBoost to
analyze a comprehensive dataset. This data encompasses crucial soil characteristics
measured by a 7-in-1 sensor, including the vital NPK (nitrogen, phosphorus, and
potassium) values and the soil's pH level. Additionally, the system considers historical
yields, climatic conditions (temperature, humidity and rainfall).
By analyzing these multifaceted parameters, the machine learning models identify
intricate relationships between soil composition, climate patterns, and successful crop
growth. This allows the system to recommend the most suitable crop for a specific land
area, maximizing yield potential and resource efficiency.

This project delves into the inner workings of each machine learning algorithm, evaluating
their strengths and weaknesses in the context of crop recommendation. A comparison of
their performance metrices like accuracy, interpretability, and computational efficiency is
done to determine the optimal algorithm for this application.