In this seminar, our focus will be on sampling and optimization, with a specific emphasis on the topic of federated sampling. We will delve into a paper where we apply two established compression techniques from the federated learning domain to the context of sampling. The outcome is a federated sampling algorithm featuring bidirectional compression. Our aim is to leverage this algorithm to gain deeper insights into the connections between sampling and optimization.
Texto informado pelo autor.
* Os participantes dos seminários não poderão acessar às dependências da FGV usando bermuda, chinelos, blusa modelo top ou cropped, minissaia ou camiseta regata. O uso da máscara é facultativo, porém é obrigatória a apresentação do comprovante de vacinação (físico ou digital).
Apoiadores / Parceiros / Patrocinadores
Avetik Karagulyan is a PostDoctoral fellow at KAUST in the team of professor Peter Richtárik. He defended his thesis at the Center of Research in Economics and Statistics (CREST), Paris under the supervision of professor Arnak Dalalyan. In 2018, he received his MSc Mathematics, Vision, Learning (MVA) diploma at ENS Paris-Saclay with highest honors. He graduated from Yerevan State University's faculty of Mathematics and Mechanics in 2017 with excellence. His research focuses on the study of sampling and optimization algorithms and their interconnections.