Abstract
We reformulate branch-and-bound feature selection employing L/sub /spl infin// or particular L/sub p/ metrics, as mixed-integer linear programming (MILP) problems, affording convenience of widely available MILP solvers. These formulations offer direct influence over individual pairwise interclass margins, which is useful for feature selection in multiclass settings.
Keywords
Integer programmingFeature selectionLinear programmingPairwise comparisonSelection (genetic algorithm)Feature (linguistics)Branch and priceArtificial intelligenceComputer scienceInteger (computer science)Multiclass classificationPattern recognition (psychology)Branch and boundMathematical optimizationGenetic programmingMachine learningMathematicsAlgorithmSupport vector machine
Affiliated Institutions
Related Publications
A Probabilistic Decision Model for Entity Matching in Heterogeneous Databases
In recent years, there has been a proliferation of database systems in all types of organizations. In many cases, these databases are developed in different departments and main...
An introduction to variable and feature selection
Variable and feature selection have become the focus of much research in areas of application for which datasets with tens or hundreds of thousands of variables are available. T...
Publication Info
- Year
- 2003
- Type
- article
- Volume
- 25
- Issue
- 6
- Pages
- 779-783
- Citations
- 38
- Access
- Closed
External Links
Social Impact
Altmetric
PlumX Metrics
Social media, news, blog, policy document mentions
Citation Metrics
38
OpenAlex
Cite This
Frank J. Iannarilli,
Paul A. Rubin
(2003).
Feature selection for multiclass discrimination via mixed-integer linear programming.
IEEE Transactions on Pattern Analysis and Machine Intelligence
, 25
(6)
, 779-783.
https://doi.org/10.1109/tpami.2003.1201827
Identifiers
- DOI
- 10.1109/tpami.2003.1201827