In this work, we propose an interactive platform to perform grammar-guided symbolic regression using a reinforcement learning approach from human-preference feedback. To do so, a reinforcement learning algorithm iteratively generates symbolic expressions, modeled as trajectories constrained by grammatical rules, from which a user shall elicit preferences. The interface gives the user three distinct ways of stating its preferences between multiple sampled symbolic expressions: categorizing samples, comparing pairs, and suggesting improvements to a sampled symbolic expression. Learning from preferences enables users to guide the exploration in the symbolic space toward regions that are more relevant to them. We provide a web-based interface testable on symbolic regression benchmark functions and power system data.

International audience

In this paper we deal with one of the most relevant problems in the field of data mining, the real time processing and visualization of data streams. To deal with data streams we propose a novel approach that uses a neighborhood-based clustering. Instead of processing each new element one by one, we propose to process each group of new elements simultaneously. A clustering is applied on each new group using neighborhood graphs. The obtained clusters are then used to incrementally construct a representative graph of the data stream. The data stream graph is visualized in real time with specific visualizations that reflect the processing algorithm. To validate the approach, we apply it to different data streams and we compare it with known data stream clustering approaches.