Performance Evaluation of Unsupervised Learning in Optimization of IOT- based Irrigation System

This article emphasizes on implementation of unsupervised learning methods for efficient irrigation management system based on sensor data obtained through IoT. The work focuses on temperature, humidity and soil moisture data through IoT sensors for irrigation. Different unsupervised learning techniques are implemented using MATLAB for classification of irrigational needs for the crops. As water is a limited resource, its optimization in agricultural domain is the prime reason to conduct this study. The dataset is pre-processed on basis of thresholds for temperature, humidity and soil moisture. Temperature above 30<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sup>C, humidity below 50% and soil moisture below 40% is indicative of irrigational need. Competitive Network, SOM and K-means clustering models were developed and tested, on IoT based sensor data, using MATLAB. The study exhibits unsupervised learning-based classification techniques, integrated with IoT sensor data, that majorly enhance water management in agricultural needs. The developed models can help greatly in real -time decision making, optimizing the use of water to attain sustainable agricultural requirements.