Interested in Machine Learning and its applications especially recommender systems
My research area of my PhD is the recommender systems and more specially the exploration-exploitation dilemma that occurs in it. Recommender sytems
have been already well studied and many algorithms exists to solve this problem. Although, most solution are often good only when users and items
are well known by the system. Besides, those algorithms usually perform pure exploitation strategy however it has been demonstrated that this strategy
is sub-optimal especially when new users and items keep coming, which is almost always the case in recommendation. That's why I focus on this problem
trying to develop new techniques to improve existing solution using any kind of technique : transfer learning, bandit stratey, use contextual information
smartly, reinforcement learning ... In a nutshell, my PhD topic is Novel Learning and Exploration-Exploitation techniques for Effective Recommender System
In parallel to this topic, I lead the data science team at fifty-five and solve daily real world problem for different website. Those problem can be e-commerce
ones like recommendation application of my research study or more general ones like user scoring, churn detection, media attributions, product recommendation, ...
Finally, on my free time, I also work as a data scientist for Data For Good, an association that brings together data scientist and non-profits
organisation to solve problems with strong impacts. Recently we work for the french Red Cross to help them with their distributions issues.
In my everyday life, I code in Python, R and sometimes Spark and Tensorflow and perform SQL requests on BigQuery and hive.