Document Detail


Introducing Uncertainty in Predictive Modeling - Friend or Foe?
MedLine Citation:
PMID:  23039214     Owner:  NLM     Status:  Publisher    
Abstract/OtherAbstract:
Uncertainty was introduced to chemical descriptors of 16 publicly available datasets to various degrees and in various ways in order to investigate the effect on the predictive performance of the state-of-the-art method ensembles of decision trees. A number of strategies to handle uncertainty in ensembles of decision trees were evaluated. The main conclusion of the study is that uncertainty to a large extent may be introduced in chemical descriptors without impairing the predictive performance of ensembles and without the predictive performance being significantly reduced from a practical point of view. The investigation further showed that even when distributions of uncertain values were provided, the ensembles method could generate equally effective models from single-point samples from these distributions. Hence, there seems to be no advantage in using more elaborate methods for handling uncertainty in chemical descriptors when using ensembles of decision trees as a modeling method for the considered types of introduced uncertainty.
Authors:
Ulf Norinder; Henrik Boström
Related Documents :
24696664 - Linear processes in stochastic population dynamics: theory and application to insect de...
23885044 - Accuracy of dietary reference intakes for determining energy requirements in girls.
24916124 - New clade of enigmatic early archosaurs yields insights into early pseudosuchian phylog...
24711584 - Using item and test information to optimize targeted assessments of psychological distr...
12445724 - Evaluation of cost-precision rations of different strategies for elisa measurement of s...
2205104 - Continuing periosteal apposition. i: documentation, hypotheses, and interpretation.
Publication Detail:
Type:  JOURNAL ARTICLE     Date:  2012-10-6
Journal Detail:
Title:  Journal of chemical information and modeling     Volume:  -     ISSN:  1549-960X     ISO Abbreviation:  J Chem Inf Model     Publication Date:  2012 Oct 
Date Detail:
Created Date:  2012-10-8     Completed Date:  -     Revised Date:  -    
Medline Journal Info:
Nlm Unique ID:  101230060     Medline TA:  J Chem Inf Model     Country:  -    
Other Details:
Languages:  ENG     Pagination:  -     Citation Subset:  -    
Export Citation:
APA/MLA Format     Download EndNote     Download BibTex
MeSH Terms
Descriptor/Qualifier:

From MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine


Previous Document:  Targeting combinations of liposomal drugs to both tumor vasculature cells and tumor cells for the tr...
Next Document:  Selective amine recognition driven by host-guest proton transfer and salt bridge formation.