Accession Number : ADA295617


Title :   Active Learning with Statistical Models.


Corporate Author : MASSACHUSETTS INST OF TECH CAMBRIDGE ARTIFICIAL INTELLIGENCE LAB


Personal Author(s) : Cohn, David A. ; Ghahramani, Zoubin ; Jordan, Michael I.


Full Text : https://apps.dtic.mil/dtic/tr/fulltext/u2/a295617.pdf


Report Date : JAN 1995


Pagination or Media Count : 7


Abstract : For many types of learners one can compute the statistically optimal' way to select data. We review how these techniques have been used with feedforward neural networks. We then show how the same principles may be used to select data for two alternative, statistically-based learning architectures: mixtures of Gaussians and locally weighted regression. While the techniques for neural networks are expensive and approximate, the techniques for mixtures of Gaussians and locally weighted regression are both efficient and accurate.


Descriptors :   *MATHEMATICAL MODELS , *STATISTICAL PROCESSES , *LEARNING , NEURAL NETS , REGRESSION ANALYSIS , WEIGHTING FUNCTIONS , ARTIFICIAL INTELLIGENCE , STATISTICAL ANALYSIS.


Subject Categories : STATISTICS AND PROBABILITY
      PSYCHOLOGY


Distribution Statement : APPROVED FOR PUBLIC RELEASE