Comparative Performance of Machine Learning Models in Predicting Plant Growth using IoT Data

The use of greenhouses in agriculture is greatly facilitated by the deployment of Internet of Things (IoT) technologies, where researchers test novel configurations for monitoring and controlling plant development. In this study, a dataset of conventional and IoT greenhouses in Tikrit University is utilized to predict plant growth variables using machine learning methods. The collection has 30,000 entries in all with a total of 14 characteristics. Data pre-processing included label encoding, feature scaling, and treatment for missing values. In this work, performance is compared with four machine learning modules: XGBoost Classifier - Random Forest Classifier- DNA Support Vector Machine (SVM)- Logistic Regression. The results showed that XGBoost performed better than other models. These results show how complex machine learning techniques could be employed to minutely and efficiently monitor plant growth attributes which in turn can augment the crop yield.