Juan Carlos Perdomo
PulseAugur coverage of Juan Carlos Perdomo — every cluster mentioning Juan Carlos Perdomo across labs, papers, and developer communities, ranked by signal.
-
New algorithm stabilizes performative feedback loops with minimal model deployments
Researchers have developed a new algorithmic procedure to efficiently find performatively stable models, which are crucial when algorithmic predictions influence user decisions. This method requires significantly fewer …
-
New research simplifies online learning to multicalibration reduction
Researchers have developed a new black-box reduction from online learning to online multicalibration, which simplifies achieving high-dimensional multicalibration. This method combines any no-regret learner with an expe…
-
Performative prediction algorithms shown to stabilize data feedback loops
Researchers have developed a new theoretical framework to understand the stability of online algorithms within performative prediction settings. Their work demonstrates that any no-regret algorithm deployed in these dyn…