Tomato Fruiting Quality Prediction Using Hydroponics and Machine Learning

Abstract

The tomato fruiting quality prediction using hydroponics and Machine Learning (ML) focuses on improving tomato quality under a micro-climate setting with the use of various sensors to monitor and analyze the parameters that affect the growth of tomato. This study employed various algorithms such as k-nearest neighbor (KNN), support vector machine (SVM), decision tree, linear regression, and random forest (RF) to find the most appropriate supervised ML algorithm in predicting the tomato fruiting quality. The Random Forest algorithm performs better than the other four ML algorithms at predicting the quality of tomato fruit in the microclimate setup. The RMSE of the Decision Tree is 0.089, the absolute error is 0.040, and the squared correlation is 0.675.