Beyond empirical risk minimization: performance guarantees, distribution shifts, and noise robustness
22/04/2026
The empirical risk minimization (ERM) approach for supervised learning chooses prediction rules that fit training samples and are “simple” (generalize). This approach has been the workhorse of machine learning methods and has enabled a myriad of applications. However, ERM methods strongly rely on the specific training samples available and cannot easily address scenarios affected by distribution shifts or corrupted samples. Robust risk minimization (RRM) is an alternative approach that does not aim to fit training examples and instead chooses prediction rules minimizing the maximum expected loss (risk). This talk presents a learning framework based on the generalized maximum entropy principle that leads to minimax risk classifiers (MRCs). In particular, MRCs can minimize worst-case expected 0-1 loss while providing performance guarantees, and are strongly universally consistent using feature mappings given by characteristic kernels. In addition, the methods presented can provide techniques that are effective in practical situations that defy conventional assumptions, such as scenarios affected by distribution shifts and corrupted samples.
|
TOWARDS SECURE AND SUSTAINABLE NON-TERRESTRIAL NETWORKS
20/4/2026
The satellite networks, emerging as megaconstellations, promise significant advancements to eliminating the digital divide, especially with the deployment of direct-to-cell connections on a mass scale. However, deploying such networks remains challenging, requiring innovation in architecture, interoperability, and security. This keynote will provide an accessible overview of the development of non-terrestrial networks (NTNs), emphasizing how they can be seamlessly integrated with terrestrial systems through approaches such as Open Radio Access Networks (O-RAN). The discussion will explore how resource allocation can be configured to ensure efficient operation in heterogeneous networks, while addressing the pressing issues of resilience and security in a highly interconnected space-terrestrial ecosystem. The talk will conclude by highlighting open research directions and long-term opportunities, pointing to how NTNs can evolve into secure, intelligent, and sustainable infrastructures that support inclusive global connectivity..
|