Generative Adversarial Networks - one of the most happening developments in Machine Learning through the lens of Statistics
Speaker: Swagatam Das – Kolkata, IndiaTopic(s): Artificial Intelligence, Machine Learning, Computer Vision, Natural language processing
Abstract
Generative Adversarial Networks (GANs) represent one of the most significant breakthroughs in deep learning over the past decade. These models consist of two neural networks engaged in a competitive two-player zero-sum game. In this competitive context, they aim to approximate the underlying probability distribution of high-dimensional objects, such as images, a classic problem that conventional statistical methods often struggle to address. GANs excel at generating remarkably realistic yet synthetic multimedia data, encompassing images, text, speech signals, and more, based on the patterns they learn from their training examples. This presentation will commence with a high-level introduction to GANs and then delve into the theoretical foundation that underpins them. Lastly, the talk will explore several unresolved issues pertaining to the statistical analysis of machine learning models based on GANs.About this Lecture
Number of Slides: 90Duration: 70 minutes
Languages Available: English
Last Updated:
Request this Lecture
To request this particular lecture, please complete this online form.
Request a Tour
To request a tour with this speaker, please complete this online form.
All requests will be sent to ACM headquarters for review.