Cristóbal Guzmán

Cristóbal Guzmán

Especialidad: Optimización, machine Learning, privacidad de datos.
Cristóbal Guzmán obtuvo un PhD in Algorithms, Combinatorics and Optimization en Georgia Institute of Technology en 2015. Es Investigador Principal en el proyecto Fondecyt 1251029, “Differentially Private Neural Network Learning with Provable Guarantees: A Sparsity Based Approach”. El objetivo de este proyecto es el diseño de mejores modelos predictivos y generativos que utilicen información sensible de individuos, protegiendo al mismo tiempo su privacidad. Específicamente, busca explorar el uso de técnicas basadas en sparsity para mejorar la precisión de los modelos de redes neuronales entrenadas bajo restricciones de preservación de la privacidad, y establecer garantías de optimalidad global para la convergencia de estos métodos.

PUBLICACIONES

Publisher: SIAM Journal on Optimization  Link>

ABSTRACT

Inspired by regularization techniques in statistics and machine learning, we study complementary composite minimization in the stochastic setting. This problem corresponds to the minimization of the sum of a (weakly) smooth function endowed with a stochastic first-order oracle and a structured uniformly convex (possibly nonsmooth and non-Lipschitz) regularization term. Despite intensive work on closely related settings, prior to our work no complexity bounds for this problem were known. We close this gap by providing novel excess risk bounds, both in expectation and with high probability. Our algorithms are nearly optimal, which we prove via novel lower complexity bounds for this class of problems. We conclude by providing numerical results comparing our methods to the state of the art.

agencia nacional de investigación y desarrollo
Edificio de Innovación UC, Piso 2
Vicuña Mackenna 4860
Macul, Chile