On accelerated first-order methods for smooth minimization

• Grant number: 20/13946-3
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: January 01, 2021
• End date: August 31, 2021
• Field of knowledge: Physical Sciences and Mathematics - Mathematics - Applied Mathematics
• Principal Investigator: Sandra Augusta Santos
• Grantee: Gabriel Rodrigues Silva Grillo
• Host Institution: Instituto de Matemática, Estatística e Computação Científica (IMECC). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil
• Associated research grant: 18/24293-0 - Computational methods in optimization, AP.TEM

Abstract

Contemporary challenges arising from the treatment of information based on huge amount of data have generated renewed interest for first-order methods. Enriched with acceleration schemes, these methods may combine low cost and efficiency for solving smooth optimization problems, unsuited to second-order approaches. Motivated by such challenges, this project addresses the study of accelerated gradient strategies, aiming to analyze possible relationships between the discrete iterative schemes and their continuous counterparts, which involve related ordinary differential equations.