Program for stream Numerical Optimization Methods with Inexact Evaluations of Objective Functions and/or Derivatives

TC-27: Optimization methods with inexact evaluations
Stream: Numerical Optimization Methods with Inexact Evaluations of Objective Functions and/or Derivatives
Room: Y313 (building Main lobby)
Chair(s): Stefania Bellavia

• A stochastic Levenberg-Marquardt method using random models with complexity results
  El Houcine Bergou, Youssef Diouane, Vyacheslav Kungurtsev, Clément Royer
• An inexact restoration trust-region method with subsampling for finite-sum minimization
  Simone Rebegoldi, Stefania Bellavia, Natasa Krejic, Benedetta Morini
• A line search based proximal stochastic gradient algorithm with dynamical variance reduction
  Giorgia Franchini
• On the role of regularization and inexactness in Interior Point Methods: a Proximal Point Perspective
  Stefano Cipolla, Jacek Gondzio

TD-27: Optimization and Deep Learning
Stream: Numerical Optimization Methods with Inexact Evaluations of Objective Functions and/or Derivatives
Room: Y313 (building Main lobby)
Chair(s): Simone Rebegoldi

• Efficiently combining first- and second-order directions in neural network training
  Marco Viola, Daniela di Serafino, Ángeles Martínez Calomardo, Mahsa Yousefi
• l1-Regularization in Portfolio Selection with Machine Learning
  Valentina De Simone, Stefania Corsaro, Zelda Marino, Salvatore scognamiglio
• A new framework combining deep learning and iterative minimization for imaging applications
  Elena Morotti, Davide Evangelista, Elena Loli Piccolomini
• Training of neural networks for monitoring electricity consumption
  Stefania Bellavia, Carlo Carcasci, Simone Rebegoldi, Marco Zini

WA-27: Recent advances in large-scale optimization
Stream: Numerical Optimization Methods with Inexact Evaluations of Objective Functions and/or Derivatives
Room: Y313 (building Main lobby)
Chair(s): Pierre-Cyril Aubin

• Kernel representation of non-negative functions with applications in non-convex optimization and beyond
  Pierre-Cyril Aubin
• Global optimisation of smooth functions using sum of squares with Reproducing Kernel Hilbert Spaces
  Ulysse Marteau-Ferey
• Projection-free methods for machine learning and data science applications
  Francesco Rinaldi, Immanuel Bomze, Damiano Zeffiro
• A local analysis for eigenvalue complementarity problems
  Mario Jelitte, Andreas Fischer, Klaus Schönefeld