COMPUTERIZED LEARNING MACHINES APPLIED TO CHEMICAL PROBLEMS - MOLECULAR FORMULA DETERMINATION FROM LOW RESOLUTION MASS SPECTROMETRY

被引：131

作者：

JURS, PC

KOWALSKI, BR

ISENHOUR, TL

机构：

[1] Department of Chemistry, University of Washington, Seattle

来源：

ANALYTICAL CHEMISTRY | 1969年 / 41卷 / 01期

关键词：

D O I：

10.1021/ac60270a002

中图分类号：

O65 [分析化学];

学科分类号：

070302 ; 081704 ;

摘要：

A computerized learning machine has been applied to the chemical problem of determining molecular formulas from low resolution mass spectra. Learning involves applying feedback to a decision process to minimize the probability of undesired responses. The basic concepts of learning machine theory are presented, along with the computer programs which, using a training set of low resolution mass spectra, empirically develop pattern classifiers to correctly determine molecular formulas. Statistically varied spectra are used to demonstrate the high reliability of the system, and some possible applications of learning machines to other aress of chemical data interpretation are discussed. © 1969, American Chemical Society. All rights reserved.

引用

页码：21 / &