Probabilistic Nearest Neighbors Classification.

Journal: Entropy (Basel, Switzerland)
Published Date:

Abstract

Analysis of the currently established Bayesian nearest neighbors classification model points to a connection between the computation of its normalizing constant and issues of NP-completeness. An alternative predictive model constructed by aggregating the predictive distributions of simpler nonlocal models is proposed, and analytic expressions for the normalizing constants of these nonlocal models are derived, ensuring polynomial time computation without approximations. Experiments with synthetic and real datasets showcase the predictive performance of the proposed predictive model.

Authors

  • Bruno Fava
    Department of Economics, Northwestern University, Evanston, IL 60208, USA.
  • Paulo C Marques F
    Insper Institute of Education and Research, Rua Quatá 300, São Paulo 04546-042, Brazil.
  • Hedibert F Lopes
    Insper Institute of Education and Research, Rua Quatá 300, São Paulo 04546-042, Brazil.

Keywords

No keywords available for this article.