Maximizing the Predictive Value of Production Rules

Abstract

A new approach to finding a solution for an important empirical learning problem is described. The problem is to find the single best production rule of a fixed length for classification. Predictive Value Maximization (PVM), a heuristic search procedure through the space of conjunctions and disjunctions of variables and their cuttoff values, is outlined. Examples are taken from laboratory medicine, where the goal is to find the best combination of tests for making a diagnosis. Resampling techniques for estimating error rates are integrated into the PVM procedure for rule induction. Excellent results for PVM are reported on data sets previously analyzed in the AI literature using alternative classification techniques. Keywords: Decision making; Artificial intelligence.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Aug 31, 1988
Accession Number
ADA200309

Entities

People

  • Casimir A. Kulikowski
  • Prasad V. Tadepalli
  • Robert S. Galen
  • Sholom M. Weiss

Organizations

  • Rutgers University–New Brunswick

Tags

Communities of Interest

  • Autonomy
  • Human Systems

DTIC Thesaurus Topics

  • Accuracy
  • Algorithms
  • Artificial Intelligence
  • Computer Science
  • Computers
  • Data Analysis
  • Data Sets
  • Databases
  • Expert Systems
  • Information Science
  • Machine Learning
  • Pattern Recognition
  • Recognition
  • Statistical Analysis
  • Statistics
  • Test Methods
  • Thyroid Diseases

Readers

  • Computational Linguistics
  • Infectious Disease/Epidemiology
  • Statistical inference.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - Machine Learning Algorithms
  • Space