Prediction of Basil height in Hydroponic Systems through Machine Learning

Sammanfattning

This thesis explores the application of machine learning techniques, a type of artificial intelligence that enables computers to learn from data, in hydroponic systems for predicting basil plant height. Hydroponics is a method of growing plants without soil, using nutrient-rich water instead. Conducted at Devoteam, an IT consulting firm in Malmö, the study involved the implementation of two hydroponic systems: Deep Water Culture and Nutrient Film Technique, focusing on monitoring and collecting data such as electrical conductivity, pH levels, water temperature. Utilizing five distinct machine learning models, namely Linear Regression, Decision Tree, Random Forest, Support Vector Regression, and K-Nearest Neighbours, we analyzed their performance in predicting basil plant height. Data were collected using a system equipped with a microcontroller unit, EC sensor, and water temperature sensor, supplemented by data from the open dataset, OpenAg. Our findings indicate that the Random Forest model consistently outperformed other models across datasets, demonstrating superior accuracy and predictive capability. This research provides insights into leveraging machine learning for optimizing hydroponic cultivation practices.