Young researcher: Oussama M’Hamdi

My name is Oussama M'Hamdi, from Tunisia. I obtained my Engineer Degree in Agricultural Production in Tunisia and a Master in Horticultural Genetics and Biotechnology in Greece. I have worked for nearly a year in a Tunisian agricultural company specializing in fruit production, overseeing peach and apple tree cultivation. Currently, I am finalizing my PhD at the Hungarian University of Agriculture and Life Sciences, funded by the Stipendium Hungaricum Scholarship, with my public defense forthcoming. My research focuses on enhancing processing tomatoes using integrative approaches. We have explored the impact of different water supplies on tomato roots using non-destructive methods, utilized machine learning to predict tomato fruit quality traits from environmental and meteorological data. Additionally, we assessed tomato plant genetic resources for Brix and lycopene content under varying Hungarian environments, considering genotype and genotype-environment interactions. My work has resulted in two impact-factor papers where I am the first author. I am passionate about statistics, mathematical models, and data analysis, and I also enjoy hands-on field and lab work. I am always eager to learn and apply new and trending methods and approaches in my research.

At the World Processing Tomato Congress, I presented two posters. The first poster, titled "Adaptive Responses of Tomato Plants to Varying Irrigation Levels: Insights into Root Development Efficiency," aimed to understand how tomato plants react to different water levels to achieve efficient water usage and promote sustainable agricultural practices. Using a camera capable of capturing root images, we periodically monitored root development and extracted various data, such as root number and root length, through specialized software.

The second poster, "Prediction of Tomato Quality Traits Utilizing Machine Learning Models," focused on developing robust predictive models for fruit quality based on environmental conditions. Utilizing a large dataset, we trained two machine learning models, XGBoost and Neural Network, to predict Brix, Lycopene, and the a/b ratio of tomato fruits. The findings highlight the importance of selecting and fine-tuning the appropriate machine learning model to enhance precision agriculture, laying the groundwork for an application that can be used daily by farmers, researchers, and breeders.

The full papers of this research will be published alongside the other research work presented during the symposium in a special edition of Acta horticulturae.

Photo: Oussama receiving the certificate from Symposium Convenor Luca Sandei