From schneide at cs.cmu.edu  Thu Nov  6 17:54:45 2014
From: schneide at cs.cmu.edu (Jeff Schneider)
Date: Thu, 06 Nov 2014 17:54:45 -0500
Subject: Thesis Proposal - Xuezhi Wang
In-Reply-To: <22F1FEC7-BA2C-4FD7-9920-B4F2E159494D@andrew.cmu.edu>
References: <22F1FEC7-BA2C-4FD7-9920-B4F2E159494D@andrew.cmu.edu>
Message-ID: <545BFC35.3030007@cs.cmu.edu>


Hi Everyone,

Tomorrow Xuezhi will do her thesis proposal!  Come out to hear about some cool 
work in transfer learning!

Jeff.



From: Deborah Cavlovich <deb at cs.cmu.edu>
Subject: Thesis Proposal - Xuezhi Wang
Date: October 27, 2014 7:52:26 PM EDT
To: cs-friends at cs.cmu.edu, cs-students at cs.cmu.edu, cs-visitors at cs.cmu.edu

CALENDAR OF EVENTS

November 7, 2014

Xuezhi Wang

1:00 PM, 4405 Gates and Hillman Centers

Thesis Proposal

Title:  Active Transfer Learning

Abstract:

Transfer learning algorithms are used when one has sufficient training data for 
one supervised learning task (the source task) but only very limited training 
data for a second task (the target task) that is similar but not identical to 
the first.  These algorithms use varying assumptions about the similarity 
between the tasks to carry information from the source to the target task. 
Common assumptions are that only certain specific marginal or conditional 
distributions have changed while all else remains the same. Moreover, not much 
work on transfer learning has considered the case when a few labels in the test 
domain are available. Alternatively, if one has only the target task, but also 
has the ability to choose a limited amount of additional training data to 
collect, then active learning algorithms are used to make choices which will 
most improve performance on the target task.  These algorithms may be combined 
into active transfer learning, but previous efforts have had to apply the two 
methods in sequence or use restrictive transfer assumptions.

This thesis focuses on active transfer learning under the model shift 
assumption. We start by proposing two transfer learning algorithms that allow 
changes in all marginal and conditional distributions but assume the changes are 
smooth in order to achieve transfer between the tasks. We then propose an active 
learning algorithm that yields a combined active transfer learning algorithm. 
Furthermore, we consider a general case where both the support and the model 
change across domains. We transform both $X$ and $Y$ by a 
parameterized-location-scale shift to achieve transfer between tasks. Previous 
work on covariate shift assumes that the support of test $P(X)$ is contained in 
the support of training $P(X)$, and target/conditional shift makes a similar 
assumption for $P(Y)$. Since we allow more flexible transformations, the 
proposed method yields better results on both synthetic data and real-world data.

In this thesis, we further propose: (1) Risk analysis on the proposed 
approaches; (2) Faster model update and extension to non-myopic active 
selection; (3) Active surveying with partial information; (4) Extension to 
multi-task learning.

Thesis Committee:
Jeff Schneider, Chair
Christos Faloutsos
Geoff Gordon
Jerry Zhu, University of Wisconsin-Madison

Thesis Summary: 
http://autonlab.org/autonweb/22755/version/1/part/5/data/proposal_wxz.pdf?branch=main&language=en