Learning from mistakes: Towards a correctable learning algorithm

Karthik Raman, Krysta M. Svore, Ran Gilad-Bachrach, Chris J.C. Burges

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Many learning algorithms generate complex models that are difficult for a human to interpret, debug, and extend. In this paper, we address this challenge by proposing a new learning paradigm called correctable learning, where the learning algorithm receives external feedback about which data examples are incorrectly learned. We define a set of metrics which measure the correctability of a learning algorithm. We then propose a simple and efficient correctable learning algorithm which learns local models for different regions of the data space. Given an incorrect example, our method samples data in the neighborhood of that example and learns a new, more correct local model over that region. Experiments over multiple classification and ranking datasets show that our correctable learning algorithm offers significant improvements over the state-of-the-art techniques.

Original languageEnglish
Title of host publicationCIKM 2012 - Proceedings of the 21st ACM International Conference on Information and Knowledge Management
Pages1930-1934
Number of pages5
DOIs
StatePublished - 2012
Externally publishedYes
Event21st ACM International Conference on Information and Knowledge Management, CIKM 2012 - Maui, HI, United States
Duration: 29 Oct 20122 Nov 2012

Publication series

NameACM International Conference Proceeding Series

Conference

Conference21st ACM International Conference on Information and Knowledge Management, CIKM 2012
Country/TerritoryUnited States
CityMaui, HI
Period29/10/122/11/12

Keywords

  • classification
  • correctable learning
  • regression

Fingerprint

Dive into the research topics of 'Learning from mistakes: Towards a correctable learning algorithm'. Together they form a unique fingerprint.

Cite this