I defended my PhD in Mathematics and Computer Science in June 2017
Member of LEAR and BiPoP Teams, INRIA, Grenoble University
LinkedIn
federicopierucci.eu
Research
During my PhD I worked with my supervisors Anatoli Iouditski (director), Zaid Harchaoui (co-encadrant), Jerome Malick (co-encadrant) on on large-scale non-smooth optimization for machine learning.
Slides of my PhD defense
Interests
- Convex Optimization
- Machine Learning
- Computer Vision
In particular
- Smoothing of nonsmooth loss functions
- Trace norm regularized optimization
- Collaborative filtering
- Multi class classification
- Frank-Wolfe / conditional gradient type methods
Publications
- F. Pierucci, Z. Harchaoui, J. Malick, 2014 - A smoothing approach for composite conditional gradient with nonsmooth loss -
Conférence d'Apprentissage Automatique (Cap'2014)
BibTeX
@conference{pierucci2014,
title={A smoothing approach for composite conditional gradient with nonsmooth loss},
author={Pierucci, Federico and Harchaoui, Zaid and Malick, J{\'e}r{\^o}me},
year={2014},
journal={Conf{\'e}rence d'Apprentissage Automatique}
}
Slides
International experiences
Seminars
Workshop co-organization
Affiliations
- SAM Team - Mathematical statistics, machine learning, signal and image processing
- Laboratoire Jean Kuntzmann - The applied mathematics and computer science laboratory in Grenoble
- ED MSTII - Doctoral school for computer sciences, applied and pure mathematics in Grenoble
- Université Joseph Fourier - The science and technology University of Grenoble
Teaching
- [Fall 2012, fall 2013, fall 2014] Mathematics MAT11a 192 hours equivalent TD - Département licence sciences et technologies - Grenoble University
Professional Affiliations
- Mathematical Optimization Society (MOS)
- Society for Industrial and Applied Mathematics (SIAM)