Search

Obtaining guarantees on the convergence of the minimizers of empirical risks to the ones of the true risk is a fundamental matter in statistical learning. Instead of deriving guarantees on the usual estimation error, the goal of this paper is to provide concentration inequalities on the distance between the sets of minimizers of the risks for a broad spectrum of estimation problems. In particular, the risks are defined on metric spaces through probability measures that are also supported on metric spaces. A particular attention will therefore be given to include unbounded spaces and non-convex cost functions that might also be unbounded. This work identifies a set of assumptions allowing to describe a regime that seem to govern the concentration in many estimation problems, where the empirical minimizers are stable. This stability can then be leveraged to prove parametric concentration rates in probability and in expectation. The assumptions are verified, and the bounds showcased, on a selection of estimation problems such as barycenters on metric space with positive or negative curvature, subspaces of covariance matrices, regression problems and entropic-Wasserstein barycenters.

Random walks on networks are widely used to model stochastic processes such as search strategies, transportation problems or disease propagation. A prominent example of such process is the guiding of naive T cells by the lymph node conduits network. Here,we propose a general framework to find network heterogeneities, which we define as connectivity patterns that affect the random walk. We propose to characterize and measure these heterogeneities by detecting communities in a way that is interpretable in terms of random walk. Moreover, we use an approximation to accurately and efficiently compute these quantities on large networks. Finally, we propose an interactive data visualization platform to follow the dynamics of the random walks and their characteristics on our datasets, and a ready-to-use pipeline for other datasets upon download. By computing quantitative feature of random walk informed communities detected within the network, we show that the lymph node conduit network is spatially coherent, however, despite its quasi-regularity, contains some random walk related heterogeneities. To evaluate these characteristics, we applied the same workflow of diffusion based community detection and analysis on the LNCN and a series of generated toy networks.