| Article ID | Journal | Published Year | Pages | File Type |
|---|---|---|---|---|
| 158583 | Chemical Engineering Science | 2007 | 12 Pages |
Abstract
The analysis of the relationship between the dominant descriptor and property values for the training set and the target compound enables better understanding of the main causes for increased prediction errors. Based on this analysis, new quantitative indicators for assessing the reliability of the predicted property value are proposed, which may reduce considerably the uncertainty associated with property prediction.
Keywords
Related Topics
Physical Sciences and Engineering
Chemical Engineering
Chemical Engineering (General)
Authors
Mordechai Shacham, Olaf Kahrs, Georgi St. Cholakov, Roumiana P. Stateva, Wolfgang Marquardt, Neima Brauner,