Title :
Active machine learning using adaptive set estimation
Author :
Joachim, D. ; Deller, J.R., Jr.
Author_Institution :
Dept. of Electr. Engr. & Comput. Sci., Tulane Univ., New Orleans, LA, USA
Abstract :
The class of set-theoretic identification and filtering methods known as the optimal bounding ellipsoid (OBE) algorithms offer significant advantages in active machine learning tasks. Benefits include adaptive and intelligent classification over large data sets, elimination of redundancy and outliers, and automated assessment of innovation in observations. This paper presents formal links between the OBE algorithms and active machine learning solutions.
Keywords :
adaptive estimation; filtering theory; learning (artificial intelligence); set theory; active machine learning; adaptive classification; adaptive set estimation; automated innovation assessment; filtering method; intelligent classification; optimal bounding ellipsoid algorithm; outlier elimination; redundancy elimination; set-theoretic identification; Aggregates; Covariance matrix; Ellipsoids; Learning systems; Machine learning; Machine learning algorithms; Particle measurements; Recursive estimation; State estimation; Strips;
Conference_Titel :
Signals, Systems and Computers, 2004. Conference Record of the Thirty-Seventh Asilomar Conference on
Print_ISBN :
0-7803-8104-1
DOI :
10.1109/ACSSC.2003.1291980
