Modern data analysis methods are expected to handle massive amounts of data that are being collected in a variety of domains, including the biological/biomedical, social, economic, artistic, and cultural domains. This seminar will provide an overview of the advances that have been made in the last decade in machine-learning and automatic data-mining approaches for dealing with the challenges that arise in such settings. This year, the seminar will focus on a broad scope of data analysis tasks, with emphasis on deep learning approaches towards them. Additional approaches, such as kernel methods and dictionary learning, will be included as well to provide a wide perspective on the field.
The seminar will be based on student presentations and discussions of recent prominent publications from leading journals in machine learning (e.g., MLJ, JMLR, TPAMI, and DMKD), computational sciences (e.g., Science, Nature, and PNAS journals), and conferences (e.g., ICML, NIPS, ICLR, and SigKDD/SigMOD). The grades in the seminar will be based on the quality of these presentations (including submitted abstracts and slides), discussions in class, and written summaries.
Class sessions: Wednesdays 2:45-5:00, AKW 307 (even though the course is officially slotted as 2:30-5:15).
First class will be on Wednesday, January 24th.
|Jan 24||Introduction to the course||Guy Wolf & Smita Krishnaswamy|
|Mar 14||Spring Break||--|
|Mar 21||Spring Break||--|