Danh Mục Sự Kiện

Xemina “DC approximation approaches for sparse optimization”

12/30/2019 11:58:48 AM

Ngày 31/12/2019, Bộ môn Khoa học tự nhiên và Công nghệ tổ chức xemina “DC approximation approaches for sparse optimization”.

Thời gian: 14h00 – 16h00, ngày 31/12/2019

Địa điểm: ISpace, Nhà C, Làng sinh viên HACINCO, 79 Nguỵ Như Kon Tum, Thanh Xuân, Hà Nội

Diễn giả: TS. Võ Xuân Thanh

Nội dung:
Sparse learning is an important topic in machine learning and data mining due to its wide applications in computer vision, image processing, bioinformatics, etc. Direct formulations of sparse learning problems involve the zero-norm in objective or constraints making them hard to optimize in general. Fortunately, by using suitable approximations, these problems can be cast in DC (Difference of Convex functions) frameworks which can be solved efficiently by DCA (DC Algorithm). In this talk, we will explore some applications of sparse learning and DC approximation approaches for sparse optimization.

 

Lên đầu trangLên đầu trang