Structure-Reactivity Correlations for Organic Molecules by Pattern Recognition.

Abstract

Pattern recognition is used to predict the relative reactivity of a molecule from its structural features. For this preliminary study the hydrolysis of substituted benzoic acids was selected as the model reaction system. Nine structurel features of seventy-nine ortho, meta, or para substituted benzoic acids were input to a master program of pattern recognition methods. These methods were 96.2% accurate in predicting the relative reactivity of the substituted acids. The features that describe the electronic structure of the substituent and its position on the aromatic ring were found to be more important for predicting the rate than the features that describe the steric size of the substituent. (Author)

Document Details

Document Type
Technical Report
Publication Date
Aug 01, 1974
Accession Number
AD0785913

Entities

People

  • B. R. Kowalski
  • J. R. Koskinen

Organizations

  • University of Washington

Tags

DTIC Thesaurus Topics

  • Benzoic Acids
  • Hydrolysis
  • Identification
  • Molecules
  • Pattern Recognition
  • Reactivities
  • Recognition

Fields of Study

  • Chemistry

Readers

  • Computer Vision.
  • Organic Chemistry

Technology Areas

  • AI & ML
  • AI & ML - Neural Networks
  • Microelectronics