Exercise Habits and Social Contagion Patterns
University of Cyprus (UCY), Cyprus
Dr. Christos Nicolaides / Dr. George Pallis
Habits, Information Theoretic Measures, Causal Inference
John Abied Hatem received his B.S. degree in computer science from Haigazian University, Beirut, Lebanon, in 2016. By the time of his graduation, he had already started working as an iOS developer. In 2016, John started his graduate studies at the American University of Beirut (AUB) where he was awarded a full graduate scholarship. In 2019, he obtained his M.S. degree in computer science from AUB. His thesis, titled “Predictive Resource Management using Deep Learning in Next Generation Passive Optical Networks”, proposes a novel dynamic bandwidth allocation approach, which employs deep learning to predict bandwidth demands of end-users to increase the utilization of Passive Optical Networks.
During his graduate studies, he worked in the department of computer science at AUB as a research assistant for different projects under the supervision of Prof. Ahmad Dhaini, Prof. Shady Elbassouni, and Prof. Haidar Safa. He also worked as a teaching assistant for several courses.
As of November 2019, he is an Early Stage Researcher for the Marie-Curie Initial Training Network “RAIS” and will be enrolled as a Ph.D. student at the University of Cyprus.
Deep Learning, Data Science, Passive Optical Networks, Dynamic Bandwidth Allocation, IoT, RPL.
My current research activity within RAIS can be summarized by this intriguing question: How can we measure something as abstract as exercise habits from observational datasets? Well, the first step will be building a solid understanding of exercise habits in particular and habits in general. Based on this understanding, an in-depth investigation of a large observational dataset will be performed to identify behavioral attributes that have habitual nature in such datasets. Next, will be developing measures capable of quantifying habitual behavior from real observational datasets. Finally, these measures will be used to predict long-term exercise behavior and future retention of habits. In short, my work is a fascinating intertwining of the disciplines of psychology, information theory, data science, and machine learning. It will provide measures to quantify habits from increasingly available observational exercise data and will offer insights into how habits shape our behaviors.
 J. A. Hatem, A. R. Dhaini, and S. Elbassuoni, “Deep Learning-Based Dynamic Bandwidth Allocation for Future Optical Access Networks,” IEEE Access, vol. 7, pp. 97307–97318, 2019.
 J. A. Hatem, H. Safa, and W. El-Hajj, “Enhancing routing protocol for low power and lossy networks,” in Proceedings of the 13th International Wireless Communications and Mobile Computing Conference (IWCMC). IEEE, 2017, pp. 753–758.