Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

Semantic Feature Construction

Fenner, Mark E (2008) Semantic Feature Construction. Doctoral Dissertation, University of Pittsburgh. (Unpublished)

Primary Text

Download (2MB) | Preview


An effective set of features is integral to the success of machine learning algorithms. Semantic feature construction is the knowledge-driven manipulation of the propositional descriptor space of a set of examples for use in a learning algorithm. Two important sources of semanticsfor feature construction are the semantic type (and associated semantic properties) and the semantic class of features. These semantics canbe captured in a knowledge base and utilized to constrain search through the space of constructed features. This dissertation presents a systemthat captures semantic feature construction knowledge and implements a search algorithm that respects that knowledge. Results are presentedfor different combinations of features generated from different successor functions used in search. These results are compiled over many learning problems and several learning algorithms. Other results are also presentedfor different levels of detail in semantic knowledge. Generally, semantics are an effective guide in the space of constructed features.


Social Networking:
Share |


Item Type: University of Pittsburgh ETD
Status: Unpublished
CreatorsEmailPitt UsernameORCID
Fenner, Mark
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Committee ChairBuchanan, Bruce Gbuchanan@cs.pitt.eduBBUCHANA
Committee MemberWiebe, Janwiebe@cs.pitt.eduJMW106
Committee MemberHauskrecht, Milosmilos@cs.pitt.eduMILOS
Committee MemberIyengar, Satishsi@stat.pitt.eduSSI
Date: 24 January 2008
Date Type: Completion
Defense Date: 14 June 2007
Approval Date: 24 January 2008
Submission Date: 6 August 2007
Access Restriction: No restriction; Release the ETD for access worldwide immediately.
Institution: University of Pittsburgh
Schools and Programs: Dietrich School of Arts and Sciences > Computer Science
Degree: PhD - Doctor of Philosophy
Thesis Type: Doctoral Dissertation
Refereed: Yes
Uncontrolled Keywords: constructive induction; domain knowledge
Other ID:, etd-08062007-175343
Date Deposited: 10 Nov 2011 19:57
Last Modified: 15 Nov 2016 13:48


Monthly Views for the past 3 years

Plum Analytics

Actions (login required)

View Item View Item