Neural Network-Graph Theory Approach to the Prediction of the Physical Properties of Organic Compounds

Andrei A. Gakh; Elena G. Gakh; Bobby G. Sumpter; Donald W. Noid

doi:10.1021/ci00020a017

Neural Network-Graph Theory Approach to the Prediction of the Physical Properties of Organic Compounds

Andrei A. Gakh
, Elena G. Gakh
, Bobby G. Sumpter
, Donald W. Noid

Center for Nanophase Matls Sciences

Research output: Contribution to journal › Article › peer-review

95 Scopus citations

Abstract

A new computational scheme is developed to predict physical properties of organic compounds on the basis of their molecular structure. The method uses graph theory to encode the structural information which is the numerical input for a neural network. Calculated results for a series of saturated hydrocarbons demonstrate average accuracies of 1–2% with maximum deviations of 12–14%.

Original language	English
Pages (from-to)	832-839
Number of pages	8
Journal	Journal of Chemical Information and Computer Sciences
Volume	34
Issue number	4
DOIs	https://doi.org/10.1021/ci00020a017
State	Published - Jul 1 1994

Access to Document

10.1021/ci00020a017

Cite this

@article{cbe683cc1a734f57bffc7a4289682494,

title = "Neural Network-Graph Theory Approach to the Prediction of the Physical Properties of Organic Compounds",

abstract = "A new computational scheme is developed to predict physical properties of organic compounds on the basis of their molecular structure. The method uses graph theory to encode the structural information which is the numerical input for a neural network. Calculated results for a series of saturated hydrocarbons demonstrate average accuracies of 1–2\% with maximum deviations of 12–14\%.",

author = "Gakh, \{Andrei A.\} and Gakh, \{Elena G.\} and Sumpter, \{Bobby G.\} and Noid, \{Donald W.\}",

year = "1994",

month = jul,

day = "1",

doi = "10.1021/ci00020a017",

language = "English",

volume = "34",

pages = "832--839",

journal = "Journal of Chemical Information and Computer Sciences",

issn = "0095-2338",

number = "4",

}

TY - JOUR

T1 - Neural Network-Graph Theory Approach to the Prediction of the Physical Properties of Organic Compounds

AU - Gakh, Andrei A.

AU - Gakh, Elena G.

AU - Sumpter, Bobby G.

AU - Noid, Donald W.

PY - 1994/7/1

Y1 - 1994/7/1

N2 - A new computational scheme is developed to predict physical properties of organic compounds on the basis of their molecular structure. The method uses graph theory to encode the structural information which is the numerical input for a neural network. Calculated results for a series of saturated hydrocarbons demonstrate average accuracies of 1–2% with maximum deviations of 12–14%.

AB - A new computational scheme is developed to predict physical properties of organic compounds on the basis of their molecular structure. The method uses graph theory to encode the structural information which is the numerical input for a neural network. Calculated results for a series of saturated hydrocarbons demonstrate average accuracies of 1–2% with maximum deviations of 12–14%.

UR - https://www.scopus.com/pages/publications/0028465825

U2 - 10.1021/ci00020a017

DO - 10.1021/ci00020a017

M3 - Article

AN - SCOPUS:0028465825

SN - 0095-2338

VL - 34

SP - 832

EP - 839

JO - Journal of Chemical Information and Computer Sciences

JF - Journal of Chemical Information and Computer Sciences

IS - 4

ER -

Neural Network-Graph Theory Approach to the Prediction of the Physical Properties of Organic Compounds

Abstract

Access to Document

Other files and links

Fingerprint

Cite this