A mixture model framework for class discovery and outlier detection in mixed labeled/unlabeled data sets

David Jonathan Miller, John Browning

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Several authors have addressed learning a classifier given a mixed labeled/unlabeled training set. These works assume each unlabeled sample originates from one of the (known) classes. Here, we consider the scenario in which unlabeled points may belong either to known/predeAned or to heretofore undiscovered classes. There are several practical situations where such data may arise. We earlier proposed a novel statistical mixture model to flt this mixed data. Here we review this method and also introduce an alternative model. Our fundamental strategy is to view as observed data not only the feature vector and the class label, but also the fact of label presence/ahsence for each point. Two types of mixture components are posited to explain label presence/absence. "Predefined" components generate both labeled and unlabeled points and assume labels are missing at random. These components represent the known classes. "Non-predeAned" components only generate unlabeled points-thus, in localized regions, they capture data subsets that are ezclusively unlabeled. Such subsets may represent an outlier distribution, or new classes. The components' predeflnedlnonpredefined natures are data-driven, learned along with the other parameters via an algorithm based on expectation-maximization (EM). There are three natural applications: 1) robust classifier design, given a mixed training set with outliers; 2) classiflcation with rejections; 3) identitication of the unlabeled points (and their representative components) that originate from unknown classes, i.e. new class discovery. The effectiveness of our models in discovering purely unlabeled data components (potential new classes) is evaluated both on synthetic and real data sets. Although each of our models has its own advantages, our original model is found to achieve the best class discovery results.

Original languageEnglish (US)
Title of host publication2003 IEEE 13th Workshop on Neural Networks for Signal Processing, NNSP 2003
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages489-498
Number of pages10
ISBN (Electronic)0780381777
DOIs
StatePublished - 2003
Event13th IEEE Workshop on Neural Networks for Signal Processing, NNSP 2003 - Toulouse, France
Duration: Sep 17 2003Sep 19 2003

Publication series

NameNeural Networks for Signal Processing - Proceedings of the IEEE Workshop
Volume2003-January

Other

Other13th IEEE Workshop on Neural Networks for Signal Processing, NNSP 2003
Country/TerritoryFrance
CityToulouse
Period9/17/039/19/03

All Science Journal Classification (ASJC) codes

  • Electrical and Electronic Engineering
  • Artificial Intelligence
  • Software
  • Computer Networks and Communications
  • Signal Processing

Fingerprint

Dive into the research topics of 'A mixture model framework for class discovery and outlier detection in mixed labeled/unlabeled data sets'. Together they form a unique fingerprint.

Cite this