web-mining agents: transfer learning tradaboostmoeller/lectures/ws-15-16/data...web-mining agents:...
TRANSCRIPT
![Page 1: Web-Mining Agents: Transfer Learning TrAdaBoostmoeller/Lectures/WS-15-16/Data...Web-Mining Agents: Transfer Learning TrAdaBoost ... Next in Web-Mining Agents: ... Title: 12-Transfer-Learning-TrAdaBoost.ppt](https://reader034.vdocuments.us/reader034/viewer/2022051723/5aac76577f8b9a59658d0d08/html5/thumbnails/1.jpg)
Web-Mining Agents: Transfer Learning
TrAdaBoost
R. Möller Institute of Information Systems
University of Lübeck
![Page 2: Web-Mining Agents: Transfer Learning TrAdaBoostmoeller/Lectures/WS-15-16/Data...Web-Mining Agents: Transfer Learning TrAdaBoost ... Next in Web-Mining Agents: ... Title: 12-Transfer-Learning-TrAdaBoost.ppt](https://reader034.vdocuments.us/reader034/viewer/2022051723/5aac76577f8b9a59658d0d08/html5/thumbnails/2.jpg)
b y : H A I T H A M B O U A M M A R
M A A S T R I C H T U N I V E R S I T Y
Based on an excerpt of: Transfer for Supervised
Learning Tasks
![Page 3: Web-Mining Agents: Transfer Learning TrAdaBoostmoeller/Lectures/WS-15-16/Data...Web-Mining Agents: Transfer Learning TrAdaBoost ... Next in Web-Mining Agents: ... Title: 12-Transfer-Learning-TrAdaBoost.ppt](https://reader034.vdocuments.us/reader034/viewer/2022051723/5aac76577f8b9a59658d0d08/html5/thumbnails/3.jpg)
Traditional Machine Learning vs. Transfer
Source Task
Knowledge
Target Task
Learning System
Different Tasks
Learning System
Learning System
Learning System
Traditional Machine Learning Transfer Learning
![Page 4: Web-Mining Agents: Transfer Learning TrAdaBoostmoeller/Lectures/WS-15-16/Data...Web-Mining Agents: Transfer Learning TrAdaBoost ... Next in Web-Mining Agents: ... Title: 12-Transfer-Learning-TrAdaBoost.ppt](https://reader034.vdocuments.us/reader034/viewer/2022051723/5aac76577f8b9a59658d0d08/html5/thumbnails/4.jpg)
Transfer Learning Definition
� Notation: ¡ Domain :
÷ Feature Space: ÷ Marginal Probability Distribution:
¢ with
¡ Given a domain then a task is :
DX
P (X)X = {x1, . . . , xn} � X
T = {Y, f(.)}
![Page 5: Web-Mining Agents: Transfer Learning TrAdaBoostmoeller/Lectures/WS-15-16/Data...Web-Mining Agents: Transfer Learning TrAdaBoost ... Next in Web-Mining Agents: ... Title: 12-Transfer-Learning-TrAdaBoost.ppt](https://reader034.vdocuments.us/reader034/viewer/2022051723/5aac76577f8b9a59658d0d08/html5/thumbnails/5.jpg)
Transfer Learning Definition
Given a source domain and source learning task, a target domain and a target learning task, transfer learning aims to help improve the learning of the target predictive function using the source knowledge, where
or Ds �= DT Ts �= TT
![Page 6: Web-Mining Agents: Transfer Learning TrAdaBoostmoeller/Lectures/WS-15-16/Data...Web-Mining Agents: Transfer Learning TrAdaBoost ... Next in Web-Mining Agents: ... Title: 12-Transfer-Learning-TrAdaBoost.ppt](https://reader034.vdocuments.us/reader034/viewer/2022051723/5aac76577f8b9a59658d0d08/html5/thumbnails/6.jpg)
Transfer Definition
� Therefore, if either :
XS �= XT PS(X) �= PT (X)
YS �= YT P (YS |XS) �= P (YT |XT )
Domain Differences
Task Differences
![Page 7: Web-Mining Agents: Transfer Learning TrAdaBoostmoeller/Lectures/WS-15-16/Data...Web-Mining Agents: Transfer Learning TrAdaBoost ... Next in Web-Mining Agents: ... Title: 12-Transfer-Learning-TrAdaBoost.ppt](https://reader034.vdocuments.us/reader034/viewer/2022051723/5aac76577f8b9a59658d0d08/html5/thumbnails/7.jpg)
Questions to answer when transferring
![Page 8: Web-Mining Agents: Transfer Learning TrAdaBoostmoeller/Lectures/WS-15-16/Data...Web-Mining Agents: Transfer Learning TrAdaBoost ... Next in Web-Mining Agents: ... Title: 12-Transfer-Learning-TrAdaBoost.ppt](https://reader034.vdocuments.us/reader034/viewer/2022051723/5aac76577f8b9a59658d0d08/html5/thumbnails/8.jpg)
Algorithms: TrAdaBoost
� Assumptions: ¡ Source and Target task have same feature space:
¡ Marginal distributions are different: XS = XT
PS(X) 6= PT (X)
Not all source data might be helpful !
![Page 9: Web-Mining Agents: Transfer Learning TrAdaBoostmoeller/Lectures/WS-15-16/Data...Web-Mining Agents: Transfer Learning TrAdaBoost ... Next in Web-Mining Agents: ... Title: 12-Transfer-Learning-TrAdaBoost.ppt](https://reader034.vdocuments.us/reader034/viewer/2022051723/5aac76577f8b9a59658d0d08/html5/thumbnails/9.jpg)
Algorithm: TrAdaBoost
� Idea: ¡ Iteratively reweight source samples such that:
÷ reduce effect of “bad” source instances ÷ encourage effect of “good” source instances
� Requires: ¡ Source task labeled data set ¡ Very small Target task labeled data set ¡ Unlabeled Target data set ¡ Base Learner
![Page 10: Web-Mining Agents: Transfer Learning TrAdaBoostmoeller/Lectures/WS-15-16/Data...Web-Mining Agents: Transfer Learning TrAdaBoost ... Next in Web-Mining Agents: ... Title: 12-Transfer-Learning-TrAdaBoost.ppt](https://reader034.vdocuments.us/reader034/viewer/2022051723/5aac76577f8b9a59658d0d08/html5/thumbnails/10.jpg)
Algorithm: TrAdaBoost
Weights Initialization
Hypothesis Learning and error calculation
Weights Update
![Page 11: Web-Mining Agents: Transfer Learning TrAdaBoostmoeller/Lectures/WS-15-16/Data...Web-Mining Agents: Transfer Learning TrAdaBoost ... Next in Web-Mining Agents: ... Title: 12-Transfer-Learning-TrAdaBoost.ppt](https://reader034.vdocuments.us/reader034/viewer/2022051723/5aac76577f8b9a59658d0d08/html5/thumbnails/11.jpg)
Algorithms: Self-Taught Learning
![Page 12: Web-Mining Agents: Transfer Learning TrAdaBoostmoeller/Lectures/WS-15-16/Data...Web-Mining Agents: Transfer Learning TrAdaBoost ... Next in Web-Mining Agents: ... Title: 12-Transfer-Learning-TrAdaBoost.ppt](https://reader034.vdocuments.us/reader034/viewer/2022051723/5aac76577f8b9a59658d0d08/html5/thumbnails/12.jpg)
Algorithms: Self-Taught Learning
� Assumptions: ¡ Source and Target task have different feature space:
¡ Marginal distributions are different:
¡ Label Space is different:
PS(X) 6= PT (X)
YS �= YT
XS �= XT
![Page 13: Web-Mining Agents: Transfer Learning TrAdaBoostmoeller/Lectures/WS-15-16/Data...Web-Mining Agents: Transfer Learning TrAdaBoost ... Next in Web-Mining Agents: ... Title: 12-Transfer-Learning-TrAdaBoost.ppt](https://reader034.vdocuments.us/reader034/viewer/2022051723/5aac76577f8b9a59658d0d08/html5/thumbnails/13.jpg)
Algorithms: Self-Taught Learning
� Framework: ¡ Source Unlabeled data set:
¡ Target Labeled data set:
DS = {(x(i)s )}mi=1
DT = {(x(j)T , y
(j)T )}nj=1 with n <<< m
Build classifier for cars and Motorbikes
![Page 14: Web-Mining Agents: Transfer Learning TrAdaBoostmoeller/Lectures/WS-15-16/Data...Web-Mining Agents: Transfer Learning TrAdaBoost ... Next in Web-Mining Agents: ... Title: 12-Transfer-Learning-TrAdaBoost.ppt](https://reader034.vdocuments.us/reader034/viewer/2022051723/5aac76577f8b9a59658d0d08/html5/thumbnails/14.jpg)
Algorithms: Self-Taught Learning
� Step One: Discover high level features from Source data by
Regularization Term Re-construction Error
Constraint on the Bases
minb,a
mX
i=1
||x(i)s �
X
k
a(k)si bk||22 + �||asi ||1
s.t. ||bk|| ⇥ 1, ⌅k ⇤ {1, . . . ,Ks}
![Page 15: Web-Mining Agents: Transfer Learning TrAdaBoostmoeller/Lectures/WS-15-16/Data...Web-Mining Agents: Transfer Learning TrAdaBoost ... Next in Web-Mining Agents: ... Title: 12-Transfer-Learning-TrAdaBoost.ppt](https://reader034.vdocuments.us/reader034/viewer/2022051723/5aac76577f8b9a59658d0d08/html5/thumbnails/15.jpg)
Algorithm: Self-Taught Learning
Unlabeled Data Set
minb,a
mXi=1
||x(i)s�X
k
a(k)s ib k||
22+�||a
s i|| 1
s.t. ||b
k|| ⇥
1,⌅k
⇤ {1, .
. . ,K s}
![Page 16: Web-Mining Agents: Transfer Learning TrAdaBoostmoeller/Lectures/WS-15-16/Data...Web-Mining Agents: Transfer Learning TrAdaBoost ... Next in Web-Mining Agents: ... Title: 12-Transfer-Learning-TrAdaBoost.ppt](https://reader034.vdocuments.us/reader034/viewer/2022051723/5aac76577f8b9a59658d0d08/html5/thumbnails/16.jpg)
Algorithm: Self-Taught Learning
� Step Two: Project target data onto the attained features by
a�Tj= argmin
aTj
||xTj �X
k
a(k)Tjbk||22 + �||aTj ||1
Informally, find the activations in the attained bases such that: 1. Re-construction is minimized 2. Attained vector is sparse
![Page 17: Web-Mining Agents: Transfer Learning TrAdaBoostmoeller/Lectures/WS-15-16/Data...Web-Mining Agents: Transfer Learning TrAdaBoost ... Next in Web-Mining Agents: ... Title: 12-Transfer-Learning-TrAdaBoost.ppt](https://reader034.vdocuments.us/reader034/viewer/2022051723/5aac76577f8b9a59658d0d08/html5/thumbnails/17.jpg)
Algorithms: Self-Taught Learning
a �Tj = argmina
Tj ||xT
j � Xk a (k)T
j bk || 22 + �||aT
j ||1
![Page 18: Web-Mining Agents: Transfer Learning TrAdaBoostmoeller/Lectures/WS-15-16/Data...Web-Mining Agents: Transfer Learning TrAdaBoost ... Next in Web-Mining Agents: ... Title: 12-Transfer-Learning-TrAdaBoost.ppt](https://reader034.vdocuments.us/reader034/viewer/2022051723/5aac76577f8b9a59658d0d08/html5/thumbnails/18.jpg)
Algorithms: Self-Taught Learning
� Step Three: Learn a Classifier with the new features
Target Task
Source Task
Learn new features (Step 1)
Project target data (Step 2)
Learn Model (Step 3)
![Page 19: Web-Mining Agents: Transfer Learning TrAdaBoostmoeller/Lectures/WS-15-16/Data...Web-Mining Agents: Transfer Learning TrAdaBoost ... Next in Web-Mining Agents: ... Title: 12-Transfer-Learning-TrAdaBoost.ppt](https://reader034.vdocuments.us/reader034/viewer/2022051723/5aac76577f8b9a59658d0d08/html5/thumbnails/19.jpg)
Conclusions
� Transfer learning is to re-use source knowledge to help a target learner
� Transfer learning is not generalization
� TrAdaBoost transfers instances
� Self-Taught Learning transfers unlabeled features
![Page 20: Web-Mining Agents: Transfer Learning TrAdaBoostmoeller/Lectures/WS-15-16/Data...Web-Mining Agents: Transfer Learning TrAdaBoost ... Next in Web-Mining Agents: ... Title: 12-Transfer-Learning-TrAdaBoost.ppt](https://reader034.vdocuments.us/reader034/viewer/2022051723/5aac76577f8b9a59658d0d08/html5/thumbnails/20.jpg)
Next in Web-Mining Agents:
Unlabeled Features Revisited Unsupervised Learning: Clustering