Researcher in Multi-Armed Bandit Techniques for Behavioral Recommendations in Oncology

The project aims to develop and evaluate Multi-Armed Bandit (MAB) techniques for personalized behavioral recommendations in oncology. The main objective is to design adaptive algorithms that progressively learn from user interactions and identify effective, personalized recommendation strategies. Activities will include a review of the state of the art, algorithm design and implementation, and evaluation using available data or simulated environments. The project will assess the feasibility, performance, and potential applicability of the proposed approaches in oncological settings.