2024/2025
Глубинные генеративные модели
Статус:
Маго-лего
Когда читается:
2, 3 модуль
Онлайн-часы:
60
Охват аудитории:
для своего кампуса
Язык:
английский
Кредиты:
6
Контактные часы:
14
Course Syllabus
Abstract
Deep generative models are widely used in many areas of applied machine learning. In this course, we dive into modern generative model architectures and algorithms for training them. The lectures will highlight the main approaches proposed by the beginning of 2021, analyze their main advantages and disadvantages. Practical examples aim at generating images, texts and other objects using variational autoencoders (VAE), generative adversarial networks (GAN), autoregressive models, normalizing flows and other approaches. Assignments are motivated by well-known applications of generative models in science and industry. The final assignment is aligned with the everyday activity of ML researchers: reading and reproducing articles.