A0425
Title: Deep generative models and functional data prediction
Authors: Tso-Jung Yen - Academia Sinica (Taiwan) [presenting]
Abstract: A novel prediction method is presented for functional data. This method aims to deliver prediction by using samples synthesized from deep generative models. This method addressed the dimension inconsistency problem by treating functional data as set data in which different observations may have different sizes and lengths. This method is demonstrated by its application to solve problems involving real-world data.
