MIT Libraries logoDSpace@MIT

MIT
View Item 
  • DSpace@MIT Home
  • MIT Libraries
  • MIT Theses
  • Graduate Theses
  • View Item
  • DSpace@MIT Home
  • MIT Libraries
  • MIT Theses
  • Graduate Theses
  • View Item
JavaScript is disabled for your browser. Some features of this site may not work without it.

AutoFE : efficient and robust automated feature engineering

Author(s)
Song, Hyunjoon
Thumbnail
DownloadFull printable version (1.102Mb)
Alternative title
Automate feature engineering : efficient and robust automated feature engineering
Efficient and robust automated feature engineering
Other Contributors
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science.
Advisor
Samuel Madden.
Terms of use
MIT theses are protected by copyright. They may be viewed, downloaded, or printed from this source but further reproduction or distribution in any format is prohibited without written permission. http://dspace.mit.edu/handle/1721.1/7582
Metadata
Show full item record
Abstract
Feature engineering is the key to building highly successful machine learning models. We present AutoFE, a system designed to automate feature engineering. AutoFE generates a large set of new interpretable features by combining information in the original features. Given an augmented dataset, it discovers a set of features that significantly improves the performance of any traditional classification using an evolutionary algorithm. We demonstrate the effectiveness and robustness of our approach by conducting an extensive evaluation on 8 datasets and 5 different classification algorithms. We show that AutoFE can achieve an average improvement in predictive performance of 25.24% for all classification algorithms over their baseline performance obtained with the original features..
Description
Thesis: M. Eng., Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, 2018.
 
This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.
 
Cataloged from student-submitted PDF version of thesis.
 
Includes bibliographical references (pages 59-61).
 
Date issued
2018
URI
http://hdl.handle.net/1721.1/119919
Department
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Publisher
Massachusetts Institute of Technology
Keywords
Electrical Engineering and Computer Science.

Collections
  • Graduate Theses

Browse

All of DSpaceCommunities & CollectionsBy Issue DateAuthorsTitlesSubjectsThis CollectionBy Issue DateAuthorsTitlesSubjects

My Account

Login

Statistics

OA StatisticsStatistics by CountryStatistics by Department
MIT Libraries
PrivacyPermissionsAccessibilityContact us
MIT
Content created by the MIT Libraries, CC BY-NC unless otherwise noted. Notify us about copyright concerns.