ONLINE LEARNING WITH NONLINEAR MODELS
Doyen Sahoo

Speaker Event details

Doyen Sahoo
Research Fellow Candidate
School of Information Systems
Singapore Management University

Date:

Time:

Venue:








26 January 2018, Friday

10am 11am

Meeting Room 4-4, Level 4
School of Information Systems
Singapore Management University
80 Stamford Road
Singapore 178902

We look forward to seeing you at this research seminar.

Allow us to share this seminar with your co-workers

Abstract

Recent years have witnessed the success of two broad categories of machine learning algorithms: (i) Online Learning; and (ii) Learning with nonlinear models. Typical machine learning algorithms assume that the entire data is available prior to the training task. This is often not the case in the real world, where data arrives sequentially in a stream, or is too large to be stored in memory. To address these challenges, Online Learning techniques evolved as a promising solution to having highly scalable and efficient learning methods which could learn from data arriving sequentially. Further, as the real world data exhibited complex nonlinear patterns, it warranted the need for development of learning techniques that could search complex hypotheses space. Among the most notable successful methods for learning nonlinear models are kernel methods and deep neural networks. While these models enable searching complex hypothesis to learn models with a better performance, they are mostly designed for the batch setting which affects their scalability, and they also suffer from the difficulty in selecting the right hypothesis search space (e.g. which kernel to use, what architecture of neural network to use, etc.). In this work we study the intersection of both these fields, and design novel algorithms that combine the merits of both online learning and nonlinear models by proposing methods that can learn nonlinear models in an online setting. Specifically, we investigate Online Learning with both Multiple Kernels and Deep Neural Networks. We develop online multiple kernel regression for regression and time series tasks; temporal kernel descriptors to capture temporal properties of data; and cost-sensitive online multiple kernel algorithms for imbalanced data streams. We also recognize several limitations in usage of Deep Neural Networks in the online setting, and propose a novel method for depth adaptation of the Neural Networks, to enable learning Deep Neural Networks in an online fashion.

About the speaker

Doyen SAHOO obtained his PhD from School of Information Systems, Singapore Management University under the supervision of Associate Professor Steven C.H. HOI. His primary research topic is Online Learning with Nonlinear Models. He works on theoretical aspects of machine learning with focus on Online Learning, Deep Learning and Multiple Kernel Learning. He also works on applications of machine learning to portfolio optimization and cyber-security. Prior to starting PhD, Doyen completed his B.Eng. in Computer Science from Nanyang Technological University.

 

 
  This research seminar is organised as a part of the recruitment procedure for the Research Fellow position.

 
 
 
 
This research is supported by the National Research Foundation, Prime Minister's Office, Singapore under its International Research Centres in Singapore Funding Initiative.
 

If you wish to unsubscribe from the Living Analytics Research Centre mailing list, please click here
© Copyright 2016 by Singapore Management University. All Rights Reserved. In order to enhance user experience and
improve our provided services, some usage data may be collected. For more information, please refer to our Privacy Policy
     
SMU SIS LARC CMU Heinz College