Date of Award
5-15-2024
Document Type
Thesis
Degree Name
Master of Science (MS)
Department
Systems Engineering
First Advisor
Kamran Iqbal
Abstract
To optimize agricultural practices and address water scarcity, this work introduces Vertical Hydroponics and Machine Learning (ML) to smart agriculture, using data from Internet of Things (IoT) sensors to enhance crop production in low and medium-income countries, particularly Tunisia's, hydroponic farms. This research is an innovative approach to enhancing crop production within indoor vertical hydroponic systems by utilizing different ML predictive models and assessing their performance. This part was done at UALR while the data collection was done at MedTech, with 21 basil plants. This study compares Linear Regression, Long Short-Term Memory (LSTM), and Deep Neural Networks (DNN) for precision farming. Results show that the LSTM model provides high accuracy but at the cost of increased computational demand, while the DNN model balances both, and Linear Regression offers fast processing for simpler tasks. This shapes the way for an advanced decision-support platform in agriculture.
Recommended Citation
Bouzid, Emna, "Design and Implementation of a Machine Learning Predictive Model for Crop Yield in Indoor Vertical Hydroponics Farming" (2024). Theses and Dissertations. 1193.
https://research.ualr.edu/etd/1193
