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
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
收藏
页码:21 / &
相关论文
共 23 条
[1]  
BLOCK HD, 1964, COMPUTER INFORMATION, P75
[2]  
BRICK D, 1964, COMPUTER INFORMATION, P139
[3]  
CASEY R, 1966, IEEE T ELECTRONIC CO, VEC15, P91
[4]  
CRITTENDEN AL, 1968, PERSONAL COMMUNICATI
[5]  
Feigenbaum Edward A, 1963, Computers and thought
[6]   LINEAR DECISION FUNCTIONS, WITH APPLICATION TO PATTERN RECOGNITION [J].
HIGHLEYMAN, WH .
PROCEEDINGS OF THE INSTITUTE OF RADIO ENGINEERS, 1962, 50 (06) :1501-&
[7]  
KISER RW, 1968, ANAL CHEM, V40, pR273, DOI 10.1021/ac60261a007
[8]   STEPS TOWARD ARTIFICIAL INTELLIGENCE [J].
MINSKY, M .
PROCEEDINGS OF THE INSTITUTE OF RADIO ENGINEERS, 1961, 49 (01) :8-+
[9]  
MINSKY M, 1961, IRE T HUM FACT ELECT, VHFE2, P39
[10]   CHESS-PLAYING PROGRAMS AND THE PROBLEM OF COMPLEXITY [J].
NEWELL, A ;
SHAW, JC ;
SIMON, HA .
IBM JOURNAL OF RESEARCH AND DEVELOPMENT, 1958, 2 (04) :320-335