The Power of Sampling and Stacking for the PaKDD-2007 Cross-Selling Problem

Paulo J.L. Adeodato; Germano C. Vasconcelos; Adrian L. Arnaud; Rodrigo C.L.V. Cunha; Domingos S.M.P. Monteiro; Rosalvo F.O. Neto

doi:10.4018/jdwm.2008040104

To Support Customers in Easily and Affordably Obtaining the Latest Peer-Reviewed Research, Receive a 20% Discount on ALL Publications and Free Worldwide Shipping on Orders Over US$ 295
Additionally, Enjoy an Additional 5% Pre-Publication Discount on all Forthcoming Reference BooksBrowse Titles

Special Offers
- Receive Complimentary Electronic Access to the First, Second, Third, and Fourth Edition of the
  Encyclopedia of Information Science and Technology with the Purchase of the Fifth Edition
  For a limited time, receive the complimentary e-books for the first, second, third, and fourth editions with the purchase of the Encyclopedia of Information Science and Technology, Fifth Edition e-book*. Offer is only valid when purchasing the fifth edition’s hardcover + e-book or e-book only option directly through IGI Global’s Online Bookstore (www.igi-global.com/books) and is not intended for use by book distributors or wholesalers. Contact Customer Service, cust@igi-global.com, for details. Offer expires July 1, 2021.
  Learn More
  Receive a 50% Discount on All IGI Global Major Reference Works in Electronic Format through GOBI
  Until December 31, 2020, you can receive a 50% discount on all Major Reference Works (i.e. encyclopedias, handbooks of research, and research anthologies) in electronic format ordered on the IGI Global InfoSci platform through GOBI. This discount is only applicable on e-books ordered on GOBI through IGI Global's InfoSci platform and e-books are available for PDF and/or ePUB download, unlimited simultaneous users, and no DRM.
  Browse Titles
- IGI Global is Now Offering a 5% Pre-Publication Discount on
  All Reference Books Ordered Through IGI Global’s Online Bookstore*
  To support customers with accessing the latest research, IGI Global is offering a 5% pre-publication discount on all hardcover, softcover, e-books, and hardcover + e-books titles.
  *5% discount offer is eligible on hardcover, softcover, e-books, and hardcover + e-books titles and is automatically applied directly to the shopping cart. Discount offer only valid on purchases made directly through IGI Global’s Online Bookstore and offer expires 30 days after the publication’s release. This automatic discount is not intended for use by book distributors or wholesalers.
  Browse Titles
  Receive Free Shipping on Orders Over US$ 295.00
  Free shipping will be automatically applied to shopping cart orders over US$ 295.00 or more before tax, which can include a combination of print and non-shippable products (i.e. e-books, e-journals, and articles/chapters). It is only available when you order directly through IGI Global’s Online Bookstore and is only valid on standard U.S. and international shipping. There will be additional charges for express shipping and limitations may apply.
  Browse Titles
- Receive a 50% Discount on All Major Reference Works in Electronic Format
  IGI Global is offering a 50% discount on their Major Reference Works when customers order directly through the IGI Global Online Bookstore. This discount includes IGI Global's Handbooks of Research, Research Anthologies, and Encyclopedias, which are ideal for institutions looking for all-encompassing reference sources.Valid until December 31, 2020.
  View Titles
- Offering a 50% Discount on All Individual E-Journal 2021 Subscriptions
  To ensure that institutions can affordably acquire critical content, IGI Global is offering a 50% discount on all individual e-journal 2021 subscriptions through IGI Global’s Online Bookstore and preferred subscription agencies, including EBSCO Subscription Services, LM Information Delivery, Otto Harrassowitz, and more. Valid until December 31, 2020.
  Browse Journals
- Must-Have Resource! InfoSci-Knowledge Solutions Databases
  IGI Global is now offering a new collection of InfoSci-Knowledge Solutions databases, which allow institutions to affordably acquire a diverse, rich collection of peer-reviewed e-books and scholarly e-journals. Ideal for subject librarians, these databases span major subject areas including business, computer science, education, and social sciences.
  Learn More
- Create a Free IGI Global Library Account to Receive an Additional 5% Discount on All Purchases
  Exclusive benefits include one-click shopping, flexible payment options, free COUNTER 5 reports and MARC records, and a 5% discount on single all titles, as well as the award-winning InfoSci^®-Databases.
  Sign Up Now!
Books
Journals
- - Journals
  - Open Access Journals
InfoSci^®-Databases
Articles/Chapters
Publish with Us
Resources
- - Instructors
  - Course Adoption
  - Teaching Cases
  - K-12 Online Learning Collection
  - Authors and Editors
  - eEditorial Discovery^® System
  - Peer Review Process
  - Ethics and Malpractice
  - COPE Membership
  - Fair Use Policy
  - Open Access Publishing
  - Editorial Services
  - FAQ
Catalogs
About Us
Newsroom

The Power of Sampling and Stacking for the PaKDD-2007 Cross-Selling Problem

Paulo J.L. Adeodato (NeuroTech Ltd. and Federal University of Pernambuco, Brazil), Germano C. Vasconcelos (NeuroTech Ltd. and Federal University of Pernambuco, Brazil), Adrian L. Arnaud (NeuroTech Ltd. and Federal University of Pernambuco, Brazil), Rodrigo C.L.V. Cunha (NeuroTech Ltd. and Federal University of Pernambuco, Brazil), Domingos S.M.P. Monteiro (NeuroTech Ltd. and Federal University of Pernambuco, Brazil) and Rosalvo F.O. Neto (NeuroTech Ltd. and Federal University of Pernambuco, Brazil)

Source Title: International Journal of Data Warehousing and Mining (IJDWM) 4(2)

DOI: 10.4018/jdwm.2008040104

OnDemand PDF Download:

Available

$37.50

Current Special Offers

No Current Special Offers

Abstract

This article presents an efficient solution for the PAKDD-2007 Competition cross-selling problem. The solution is based on a thorough approach which involves the creation of new input variables, efficient data preparation and transformation, adequate data sampling strategy and a combination of two of the most robust modeling techniques. Due to the complexity imposed by the very small amount of examples in the target class, the approach for model robustness was to produce the median score of the 11 models developed with an adapted version of the 11-fold cross-validation process and the use of a combination of two robust techniques via stacking, the MLP neural network and the n-tuple classifier. Despite the problem complexity, the performance on the prediction data set (unlabeled samples), measured through KS2 and ROC curves was shown to be very effective and finished as the first runner-up solution of the competition.

Complete Article List

Search this Journal:

Reset

Open Access Articles

Volume 17: 4 Issues (2021): Forthcoming, Available for Pre-Order

Volume 16: 4 Issues (2020)

Volume 15: 4 Issues (2019)

Volume 14: 4 Issues (2018)

Volume 13: 4 Issues (2017)

Volume 12: 4 Issues (2016)

Volume 11: 4 Issues (2015)

Volume 10: 4 Issues (2014)

Volume 9: 4 Issues (2013)

Volume 8: 4 Issues (2012)

Volume 7: 4 Issues (2011)

Volume 6: 4 Issues (2010)

Volume 5: 4 Issues (2009)

Volume 4: 4 Issues (2008)

Volume 3: 4 Issues (2007)

Volume 2: 4 Issues (2006)

Volume 1: 4 Issues (2005)

View Complete Journal Contents Listing

MLA

APA

Chicago

Export Reference

The Power of Sampling and Stacking for the PaKDD-2007 Cross-Selling Problem

Abstract

Complete Article List